WO2009000775A1

WO2009000775A1 - Post-screening labeling of on-bead compounds

Info

Publication number: WO2009000775A1
Application number: PCT/EP2008/057878
Authority: WO
Inventors: Manfred Auer; Martin Hintersteiner; Thierry Kimmerlin
Original assignee: Novartis Ag
Priority date: 2007-06-22
Filing date: 2008-06-20
Publication date: 2008-12-31

Abstract

The present invention relates to new on bead chemistry to produce compound libraries used in improved screening methods, allowing the direct solution confirmation of on-bead screening hits without compound decoding and resynthesis. More specifically, the invention relates to on-bead compounds having the formula B - L - C - S, wherein - represents a bond or any stable covalent spacer, B is a solid support to which a plurality of L-C-S molecules are covalently attached, L is a cleavable linker for covalently linking C to the solid support, C is a covalent linking core comprising at least one protected labeling functional group for covalently linking of a label or a detection molecule, S is a spacer comprising at least one protected or unprotected functional group for on-bead compound derivatization, and wherein L and C are selected so that (i) the linker L and the protected labeling functional group are not affected by side chain deprotection during on-bead compound derivatization, (ii) the linker L and any synthetic molecule attached to S are not affected by deprotection of labeling functional group and labeling reaction step, (iii) covalent linkage of the label or the detection molecule or any synthetic molecule attached to S are not affected by cleavage of the linker L.

Description

POST-SCREENING LABELING OF ON-BEAD COMPOUNDS

The present invention relates to new on bead chemistry to produce compound libraries used in improved screening methods, allowing the direct solution confirmation of on-bead screening hits without compound decoding and resynthesis. More specifically, the invention relates to on-bead compounds having the formula (I) B - L - C - S

(I) wherein

- represents a bond or any stable covalent spacer,

B is a solid support to which a plurality of L-C-S molecules are covalently attached, L is a cleavable linker for covalently linking C to the solid support, C is a covalent linking core comprising at least one protected labeling functional group for covalently linking of a label or a detection molecule,

S is a spacer comprising at least one protected or unprotected functional group for on- bead compound derivatization,

and wherein L and C are selected so that

(i) the linker L and the protected labeling functional group are not affected by side chain deprotection during on-bead compound derivatization,

(ii) the linker L and any synthetic molecule attached to S are not affected by deprotection of labeling functional group and labeling reaction step,

(iii) covalent linkage of the label or the detection molecule or any synthetic molecule attached to S are not affected by cleavage of the linker L.

BACKGROUND

The modem drug discovery process - despite all recent technological advances - is perceived as an increasingly cost-intensive, lengthy and complex multi-step process. The development of all drugs starts from chemical substances which are synthesized individually or in libraries and stored in large compound archives. The development of reliable combinatorial and parallel synthesis strategies during the 1990ies made huge compound collections accessible, supplementing historical compound archives. The amount of substance needed for screening was considerably reduced by the miniaturization of assay volumes down to the microliter scale, and the process automatization in large screening factories became reality.

Despite the progress made, the still unchanged classical concept of purifying several milligrams of low molecular weight compounds and building up of large solution- or solid compound-archives for testing in high-throughput screening is associated with extensive storage, liquid handling and maintenance costs.

Additionally, treating high-throughput screening units as "screening factories" and hence carrying out the drug discovery process as some kind of craftsmanship does hardly do justice to the underlying scientific problems. Therefore a great deal of research in the academic world, as well as throughout the industry, has been carried out in search for novel screening concepts (Blackwell, et al 2001 ; Bradner, et al 2006; demons, et al 2001 ; Metzger, et al 2006; Muckenschnabel, et al 2004; Urbina, et al 2006; Zehender, et al 2004).

If it were possible to test compounds for their biological activity or binding affinity to target proteins directly at the site of synthesis - on-bead - only interesting substances would be purified and characterized. As a result only ~ 10-100 ng of each individual compound on- bead would be needed. Storing and handling logistics of on-bead archives would be much simpler, with reducing costs and increasing screening speed.

The basic idea of screening ligands directly on-bead appeared in the literature as early as 1992 (Lam, et al 1991). Many steps have been taken since then (Meisner, et al 2004; UhI, et al 2002; Lam, et al 2003; Lam, Lebl and Krchnak 1997; Lehman, et al 2006; Meldal 2002; Paulick, et al 2006; Song, et al 2003; Wang, et al 2005a; Wang, et al 2005b; Youngquist, et al 1994) to realize and further develop this idea.

The 4 main pre-requisites for a reliable on-bead screening process are: (a) a screening method, allowing to detect specific target binding, despite the background fluorescence of beads (preferably in an automated fashion), (b) a method for the reliable retrieval of hit beads in combination with an effective structure decoding process, (c) a method to determine whether the target binding event detected on-bead also results in complex formation between ligand and receptor in homogeneous solution and, finally, (d) a close link to cellular (mechanistic) testing of hit compounds.

For the primary on-bead screening process, besides the need to deal with the possible background fluorescence problem, the bead itself as the on-bead screening compartment has specific features which can be both, advantageous and disadvantageous compared to solution screening. This somewhat "double-sided" nature comes from the high local compound concentrations on beads. Even low affinity binders, up to millimolar Kds, can easily be detected (Ying, et al 2005). This range of detectable affinities can be adjusted to some degree by increasing the stringency of the screening conditions (Reddy, Bachhawat- Sikder and Kodadek 2004). Nevertheless it might easily happen that a reproducible on-bead binding event between a fluorescently labeled target on-bead is of too low affinity to be detected with standard fluorescent detection technologies in homogeneous solution. Hit compounds below a certain affinity to a protein target will hardly ever be followed up in a chemistry program. Therefore, it is of crucial importance for an efficient screening process to have a secondary solution confirmation method available for eliminating false positives and undesirably low affinity binders. The importance of such secondary confirmation assays has recently received much attention (Kodadek and Bachhawat-Sikder 2006). Reported confirmation methods with re-synthesized material comprise biacore (Beebe, et al 2000; Sweeney, et al 2005), ITC (Alluri, et al 2003), photo-crosslinking (Amini, et al 2003; Lin and Kodadek 2005), pull down experiments from cell lysates, and fluorescence anisotropy (Liu, et al, 2005). Some of these methods are quite time-consuming and/or need a lot of material such as photoaffinity labeling and ITC, respectively. In the drug discovery context, "high throughput" methods, which allow to determine Kds of hit compounds in solution, are clearly preferable (e.g. fluorescence anisotropy titrations). However, modifying hit compounds for use in secondary assays (e.g. incorporating tethers and tagging groups) and the case-by- case choice of a certain readout are not suitable for a "high troughput screening" process.

Tagged libraries represent a generic solution for this confirmation problem and have been used in chemical genomics and proteomics (Backes, et al 2000; Khersonsky, et al 2003; Mitsopoulos, Walsh and Chang 2004; Uttamchandani, et al 2004). Besides facilitating biological activity testing in vitro, (fluorescently) tagged libraries hold great potential for studying a compound's molecular mode of action in vivo (Alexander, et al 2006; Khersonsky, et al 2003).

Such tagged libraries are described for example in WO02/36575. The compounds of the library comprise an indazole based UV-dye (AIDA) exhibiting remarkable stability during library synthesis and proving compatible with the production of more than 2 million on-bead compounds with eight different scaffolds. The AIDA dye and a five atom diaminopropane spacer were integrated between the bead and the compound of interest. Therefore, all beads contained a fluorescence marker. The problem of using fluorescence detection in both, the on-bead and off-bead phase of the screening process, was solved by means of spectral separation. AIDA stays invisible during the on-bead screening process, where blue to red emitting dyes on proteins are used to detect target protein binding to beads. After compound cleavage from bead AIDA was used as structural decoding tool, providing specific known fragmentation patterns in mass spectrometric analysis. Finally, after re-synthesis AIDA- fluorescence allowed the detection of target-compound interactions in homogeneous solution by fluorescence anistropy, fluorescence intensity detection or in context with SEC or dialysis methods. Thus AIDA works as a built-in general solution confirmation and decoding tool (Auer and Gstach 2001).

Nevertheless, using permanently fluorescently tagged on-bead libraries introduces additional prerequisites, regarding the nature and stability of the tag and, as a consequence, is associated with several disadvantages for the whole screening process.

The inherent chemical and photophysical instability of long-wavelength dyes is largely incompatible with the process of library synthesis. On the other hand, the short excitation wavelengths of the more stable UV-dyes, like the bifunctional indazole derivative AIDA (excitation at ~ 350 nm), prevent their use in a highly parallelized microtiter plate titration system with e.g. standard single molecule spectroscopy methods. As a consequence, solution measurements with Kd determinations can only be undertaken after re-synthesis of single hit-compounds on standard fluorometers. The incorporation of a tag as part of the library compounds provides an additional interaction element with target recognition potential.

Therefore there is still a need to find a secondary confirmation method of the primary on-bead hits for target binding in solution.

The present invention fulfills this need by allowing the direct solution confirmation of on-bead screening hits without compound decoding and resynthesis. More specifically, by introducing a detection molecule, such as a fluorescent dye only after synthesis and screening of the library, the present invention overcomes the drawbacks associated with fluorescently tagged libraries and their synthesis.

The method of the invention makes the whole process of on-bead screening faster, cheaper and leaner and therefore more effective. Additionally, the generic labeling methodology for combinatorial solid-phase libraries can also be a technology with a wide range of applications in the biochemical and cellular high-throughput screening with an immediate possibility to test hits in vivo.

DETAILED DESCRIPTION

The present invention generally relates to on-bead screening methods comprising the following steps:

(a) compound library synthesis on a solid support, e.g, resin beads (one-bead one- compound),

(b) on-bead screening of the compound libraries,

(c) confirmation of the binding event(s) obtained on solid support in solution, and,

(d) decoding the unique chemical structure causing the binding event to the target molecule(s).

The on-bead compound library is designed according to the principle of the invention so that the compounds can be labeled after their synthesis on-bead and their screening and isolation against a specific target. The general chemical setup enabling such strategy is summarized in Figure 1. It consists of a cleavable linkage system L for attaching the compounds to a solid support B, e.g., a resin bead, a labeling functionality, which might be protected during the synthesis and screening steps. After synthesis and screening, any protective group has to be removed prior to performing a labeling reaction on-bead. This strategy hence requires a three-fold orthogonality as side chain deprotection, labeling functionality deprotection and linker cleavage take place at three different time points.

Therefore, in one aspect, the present invention relates to a starting material for on- bead compound library synthesis, said starting material having formula (I):

B - L - C - S

(I)

wherein

- represents a bond or any stable covalent spacer,

B is a solid support to which a plurality of L-C-S molecules are covalently attached,

L is a cleavable linker for covalently linking C to the solid support, C is a covalent linking core comprising at least one protected or unprotected labeling functional group for covalent linking of a label or a detection molecule, S is a spacer comprising at least one protected or unprotected functional group for on- bead compound synthesis

and wherein L and C are selected so that

(i) the linker L and the protected labeled functional group are not affected by side chain deprotection during on-bead compound synthesis and derivatization,

(iii) covalent linkage of the label or the detection molecule or any synthetic molecule attached to S is not affected by cleavage of the linker L.

As used herein, the term "on-bead compound" refers to a compound comprising synthetic test molecules bound to a solid phase support. Preferably, one support contains many copies of a single structural species ("one-bead one-compound"). For ease of reading, the term "on- bead" will be used hereafter, but the one skilled in the art will understand that this term does not restrict the nature and the shape of the solid phase support in any way and any appropriate solid phase support can be used for making on-bead compounds.

Preferred solid supports B, useful in the screening process of the invention, satisfy the criteria of not only being suitable for organic synthesis, but also suitable screening procedures, such as "on-bead" screening as described below.

As used herein, the term "on-bead compound libraries" refers to a collection of compounds on separate phase support particles in which each separate phase support particle contains a single structural species of a synthetic test molecule. Each support contains many copies of the single structural species.

In one embodiment, the solid support B is selected among the group consisting a polymer bead, thread, pin, sheet, membrane, silicon wafer, and a grafted polymer unit. More preferably, said solid support according to the subject of the invention are resin beads, e.g., those made of resins selected among the group consisting of: functionalized base resins, polyacrylamide based polymers, PolyOxyEthylene-PolyOxyPropylene polymers (Renil and Meldal 1996; available from Versamatrix, Copenhagen, Denmark), Super Permeable Organic Combinatorial Chemistry polymers, polystyrene/polydimethylacrylamide composites (Rademann, et al 1999; available from Versamatrix, Copenhagen, Denmark), PEGA resins (Meldal 1992; commercially available from Polymer Laboratories, Amherst, Mass.), polystyrene-polyoxyethylene based supports, Tentagel (commercially available from Rapp polymere, Tubingen, Germany), glass-based supports, controlled pore glass (commercially available from VitraBio, Steinach, Germany),

The linker L is any appropriate cleavable linker, especially those selected among the ones used in combinatorial chemistry. For example, appropriate cleavable linkers include without limitation, acid labile (for example, the Rink amide as described in (Rink 1987) and traceless silyl linkers as described in (Plunkett and Ellman 1995)), base labile (for example, HMBA as described in Atherton, Logan and Sheppard 1981), photolabile (for example, 2-nitrobenzyl type as described in Holmes and Jones 1995), safety catch type linkers (for example, SCALL linker, described in (Patek and Lebl 1991), and Kenner type linkers as described in (Maclean, Hale and Chen 2001) redox-labile, as well as other specific cleavage entities (allyl, silyl, safety catch sulfonamide), and the like.

Any stable covalent spacer originating from groups known to be useful as spacers in combinatorial, peptide and oligonucleotide chemistry with the ability to covalently link B to L, L to C or C to S can be used.

S is also a stable covalent spacer further comprising a protected or unprotected functional group for on-bead compound derivatization. Preferably S comprises a protected functional group selected among the group consisting of amines, amides, alcohols, aldehydes, carboxylic acids, sulfonic acids, sulfonamides, ketones, azides, thiols, phenols, anhydrides and the like.

In a specific embodiment of a starting material of formula (I)₁ L is a base labile linker, e.g., HMBA linker and C is an alkyne-containing amino acid.

Preferably, a compound of formula (I) has the structure of formula (II)

(H)

wherein R is a protected or unprotected functional group for on-bead compound derivatization and B is a resin bead.

In such compound, hydroxymethylbenzoic acid [HMBA] is used as a cleavable linker, featuring a two-step cleavage procedure (Sheppard and Williams 1982).

In another specific embodiment of a starting material of formula (I), L is a acid labile safety catch linker and C is an alkyne-containing amino acid.

Preferably, a compound of formula (I) in this context has the structure of formula (III)

In such compound, a novel safety catch linker is used, the preparation of which is described below. In another specific embodiment of a starting material of formula (I), L is an acid labile safety catch linker and C is a terminal alkyne group.

Preferably, a compound of formula (I) in this context has the structure of formula (IV)

(IV)

wherein B is a resin bead.

In such compound, a novel safety catch linker is used, the preparation of which is described below.

The invention thus also relates to a method for synthesis of an on-bead compound library, comprising the following steps:

(i) providing a starting material of formula (I) as described above,

(ii) if necessary, deprotecting functional group of S for compound derivatization, hereafter referred to as side chain deprotection,

(iii) synthesizing test molecules attached to S, each solid support B comprising many copies of a single structural species of a test molecule.

The on-bead compound library may be synthesized using the compound of formula (I) as a starting material by known processes, for example, by parallel synthesis giving rise to small libraries (10 to 1000 members) (Hu, et al 2006), or by split/mix or split and combine methodology, as described, for example, in (Lam, et al 1991 ). The split/mix or split and combine method is a preferred method for generating a large library, due to the exponential increase in the number of varied compounds produced. The split/mix method gives rise to a one-bead-one-compound library of large size (1000 to millions of members). Side chain deprotection can be carried out on the support with appropriate amounts of TFA and additives.

The compound library members may be built up by performing all synthetic test molecule forming reactions directly on a solid phase. Alternatively, the compound library members can be prepared by linking together preformed building blocks on a solid phase, e.g., combinatorial chemistry. The resulting library members can be small organic molecules or oligomeric compounds. In both cases, the molecules contain a variety of functional groups. The functional groups can be, for example, alkynes, aldehydes, amides, amines, carbamates, carboxylates, esters, hydroxyls, ketones, thiols, ureas, and the like. The small organic molecule can belong to various classes of compounds, including but not limited to heterocycles (for example, hydantoins, benzodiazepines, pyrrolydines, isoquinolines), carbocyclic compounds, steroids, nucleotides, alkaloids, and lipids (for reviews containing examples see: (Costantino and Barlocco 2006; Horton, et al 2005).

Accordingly, in a second aspect, the invention relates to a member of an on-bead compound library, also designated, for the ease of reading, as an "on-bead compound" and represented by formula (V)

B - L - C - X

(V) wherein

- , B, L and C are as defined above for the starting material of formula (I) and,

X is any synthetic test molecule,

In a specific embodiment of an on-bead compound of formula (V), L is a base labile linker and C is an alkyne-containing amino acid.

In a more preferred embodiment, an on-bead compound of generic formula (V) has the specific formula (Vl):

(Vl) wherein B is a resin bead and X is any synthetic test molecule.

In another specific embodiment of an on-bead compound of formula (V), L is an acid sensitive safety catch linker and C is an alkyne-containing amino acid.

In a more preferred embodiment, an on-bead compound of generic formula (V) has the specific formula (VII):

(VIl) wherein B is a resin bead and X is any synthetic test molecule.

In another specific embodiment of an on-bead compound of formula (V), L is an acid sensitive safety catch linker and C is a terminal alkyne group.

In a more preferred embodiment, an on-bead compound of generic formula (V) has the specific formula (VIII):

(VIII) wherein B is a resin bead and X is any synthetic test molecule.

The invention naturally also relates to an on-bead compound library comprising a mixture of compounds of formula (V) as defined above. Preferably, each solid support carries many copies of a single structure species of a synthetic test molecule and said library comprises at least 100 different structure species of synthetic test molecules, preferably at least 10,000, and more preferably one million different structure species of synthetic test molecules.

In a preferred embodiment, said solid support B is a resin bead.

The on-bead compounds are specially designed to provide a simple and efficient solution to perform a secondary assay in solution without the need of decoding and resynthesizing the synthetic test molecule selected and isolated from the primary on-bead screening.

More specifically, according to the screening method of the invention described below, after synthesis and screening of an on-bead compound library against a target of interest, a member of the on-bead compound library carrying a synthetic molecule potentially having a functional effect upon the target of interest is isolated and a label or detection molecule is reacted by covalent linking to the core part of the isolated on-bead compound, for example, to an alkyne-containing amino acid. The resulting compound of this labeling step is hereafter referred to as a labeled on-bead compound. Said labeled on-bead compound can then be cleaved for releasing the labeled synthetic test molecule.

Therefore, in a third aspect, the invention relates to a compound of formula (IX)

S - L - C - X I LABEL

(IX)

wherein

- , B , L, C and X are as defined above for a compound of formula (V) and LABEL is a labeling or a detection molecule.

As used herein, the terms "detection molecule" and "labeling molecule" are used respectively to designate any molecule useful for detecting a binding event of a target of interest to an on-bead compound in solution, directly (because the molecule itself is detectable) or indirectly (because the molecule can be coupled to a detectable molecule).

More specifically, said LABEL can be selected among the group consisting of

(1) a spectroscopic probe such as a fluorophore, a chromophore, a magnetic probe or a contrast reagent, or also a probe useful in electron microscopy;

(2) a radioactively labeled molecule;

(3) a molecule which is one part of a specific binding pair which is capable of specifically binding to a partner, e.g., biotin, which can bind to avidin or streptavidin;

(4) an enzyme capable of catalyzing a detectable reaction;

(5) a molecule possessing a combination of any of the properties listed above.

In a specific embodiment, said LABEL is any fluorescent dye from the class of rhodamines (e.g. Tetramethylrhodamine, Sulforhodamine, Rhodamine Green), or from the Alexa series (e.g. Alexa488, Alexa647, Alexa594), or from the cyanine series (e.g. Cy3, Cy5, Cy5.5, Cy7), or from the Bodipy series (e.g. Bodipy TMR), or from the coumarin series (e.g. 7- dimethylaminocoumarin), or from any other class of fluorescent heterocycle. The LABEL may also be a fluorescent protein, such as Green fluorescent proteins or fluorescent mutants thereof. The LABEL can also be a molecule that produces chemoluminescence, such as luciferase oraequorin.

In another specific embodiment, said LABEL is an enzyme capable of catalyzing a detectable reaction, such as for example phosphatase or peroxidase. The LABEL may furthermore be a metal, for example gold. The on-bead compound of formula (V) may be labeled with the LABEL by any conventional method depending on the nature of the LABEL and the covalent core C component.

In a specific embodiment of a labeled on-bead compound of formula (IX), L is a base labile linker or a safety catch linker and C is an alkyne-containing amino acid. For synthesis of a labeled on-bead compound of formula (IX), said LABEL comprises a functional group appropriate for making covalent linkage by the "click" reaction, with an alkyne group (KoIb and Sharpless 2003), e.g., it comprises an azide group. Examples of synthesis schemes are given hereafter in the experimental part.

In a more preferred embodiment, a compound of generic formula (IX) has the specific formula (X):

(X)

wherein B is a resin bead and X is any synthetic test molecule.

In a more preferred embodiment, a compound of generic formula (IX) has the specific formula (Xl):

(Xl) wherein B is a resin bead and X is any synthetic test molecule

In a more preferred embodiment, a compound of generic formula (IX) has the specific formula (XII):

(XII) wherein B is a resin bead and X is any synthetic test molecule

In addition to providing on-bead compound libraries of a great variety of chemical structures as synthetic test molecules, and methods of synthesis thereof, the present invention further pertains to a method for identifying a test molecule that has a functional effect upon a target, the method comprising: a) contacting on-bead compound libraries of the invention with a target of interest, b) determining the functional effect of on-bead compound members upon the target, c) isolating an on-bead compound carrying test molecule having a functional effect upon the target.

As used herein, the term "target" refers to any molecule for which binding capacity is sought with a test molecule. It can be for example a protein such as a receptor or an enzyme. Preferred targets are natural receptors or enzymes involved in a biologic pathway.

As used herein, the term "functional effect" can be any effect that can be measured in conventional screening assays, either cellular or non-cellular. Preferably, said functional effect is a binding event (the target binding to the test molecule), modulation of an enzymatic activity and/or modulation of a pair-binding event, e.g., the test molecule inhibiting a ligand- receptor interaction.

A variety of suitable methods are useful for determining the target-compound functional effect.

In a specific embodiment wherein said functional effect is a target-compound binding event, target molecule bound to specific on-bead compounds can be detected directly by labeling the target with a detection probe. Any detection probe can be used, it can be identical to or different from the LABEL which can be used in the binding assay with the released test compound. Hence, in another specific embodiment of the screening method, the target is labeled with a detection probe, for example, with a fluorescent dye such as Tetramethylrhodamine, Rhodamine Green, Alexa488, Alexa594, or Atto633. In particular, when more than one target is used it is preferred that the targets are labeled with a detection probe prior to incubation with the compound library. Thus, in a specific embodiment, different targets are labeled with different detection probes, to allow primary binding assay with different targets, further confirmed in solution with individual targets. The individual targets may be labeled using different or similar detection probes. Hence, in such specific embodiments, at least 2 or more targets, such as 3, for example 4 or 5, for example in the range of 5 to 10 differentially labeled targets are used with the method of the invention. For example, after binding of the labeled target to on-bead compounds, corresponding individual beads are detected via appropriate detection methods. If fluorescently labeled targets are used, labeled beads can be detected by using standard fluorescence microscopes. The selected beads are then isolated using a commercially available beadsorter (e.g. COPAS, available from Union Biometrica, a Harvard Bioscience Company, MA₁ USA. described in (Meldal 2002)) in combination with "eye" detection and manual bead picking. Alternatively, one can use a semi-automatic bead picking device such as a device for confocal nanoscanning for on-bead screening ((Meisner, et al 2004; UhI, et al 2002)). This unique high resolution optical method allows the reliable detection and quantification of binding events in the outer ~ 1-2 μm of (e.g., TentaGel) beads with superior sensitivity and effectively suppresses the background intensity arising from the bead matrix.

The invention further relates to a method for screening on-bead compounds having a functional effect upon a target of interest and for generating corresponding labeled test molecules suitable for secondary solution assay, said method comprising the following steps further to steps a) to c) described above: d) deprotecting the labeling functional group of the covalent linking core C of the isolated on- bead compound, e) reacting a LABEL with the covalent linking core C of the selected on-bead compound to yield a labeled on-bead compound, f) cleaving the linker under appropriate conditions to release the labeled synthetic test compound from the bead, g) recovering the labeled synthetic test molecule from step (f).

At step (d), the labeling reaction enables to obtain a compound of formula (V) as described above. The labeling reaction can be performed on all or part of isolated beads from step (c).

In one specific embodiment, at step (d), a fluorescent dye is used as a LABEL and secondary binding assay of the released labeled synthetic test molecule in solution is performed by any appropriate spectroscopic method such as 2D-FIDA, FIDA, FCS, FIMDA, BIFL, cTRA, FILDA, cFLA, or any other confocal or non-confocal fluorescence spectroscopic method, probing compound/protein interactions ((Eggeling, et al 1998; Gall, et al 2002; Kask, et al 1999; Kask, et al 2000; Palo, et al 2002; Palo, et al 2000a; Palo, et al 2000b; Rothwell, et al 2003; Widengren, et al 2006)).

With the method of the invention it is possible to obtain quantitative solution affinity data for all on-bead screening hits directly, without resynthesis. Moreover, the small amount of labeled compound which can be recovered from one hit bead should be sufficient for thousands of single-molecule spectroscopic confirmation assays of, e.g., 5 μl volume. In the context of on-bead screening, this possibility to generate tens to hundreds of assay points from each hit bead opens up new routes for designing and running an efficient on-bead screening process. These new features of on-bead screening process include without limitation: a) Multiplexing by screening multiple targets simultaneously, a quantitative interaction/specificity profile is generated for each hit compound. In contrast to hitherto existing multiplexing approaches, the assignment of the respective target to a certain hit bead can be accomplished in the solution confirmation step by testing every hit compound against each target in the screen. This does not only increase the obtainable throughput but generates additional specificity information about the obtained hit compounds. Such interaction/specificity profiles are of special interest when considering biased libraries and/or related targets or whole target classes. b) Cellular confirmation - the generation of fluorescent ligands with "tuneable" fluorescence properties broadens the potential scope of the on-bead screening process. Demonstrating target/compound interactions in a cellular context represents a new level of confirmation for an in vitro screening system, the ultimate goal being to resolve the mechanistic mode of action for an identified hit compound. c) Resource efficiency and speed - compounds are decoded only after solution confirmation and ranking. This effectively reduces the number of beads to be analyzed by HPLC-MS or MALDI-MS. The attractiveness of screening compounds on-bead lies in the efficient resource management, as only very low amounts of compound and no purification are needed. By decoding and resynthesizing only proven hit compounds, activities are very efficiently shifted towards active compounds.

For example, in one specific embodiment, a portion of the released labeled candidate compound obtained at step (g) is used to confirm binding capacity to the target in solution and another portion is used for further analysis such as decoding of structure of the test molecule, e.g. by mass spectrometry decoding or cellular screening assays or cell penetration assays.

The deliverables of such an on-bead screening process are even not limited to the discovery of drug candidates; in addition, fluorescent ligands may serve as starting points in several different approaches and provide valuable tools for further experiments. FIGURE LEGENDS

Figure 1 Strategies for a generic PostSynthesis/PostScreening labeling

Figure 2 First experimentally investigated setup for PS/PS labeling : Resins 1 a,b served as the base resins for synthesis of various β-peptidic compounds (see below). The use of an α- or β³-propargylglycine adds further flexibility. The β³-propargylglycine is advantageous for incorporation into "all-β³" helix-mimetics, where it can stabilize the secondary structure and provide the derivatization site.

Figure 3 Synthesis of 3-azido-propylamido-tetramethylrhodamine (5)

Figure 4 Synthesis of three test peptides on the PS/PS labeling platform Three different β-peptides 6-8 were synthesized on TentaGel beads and used to establish a generic PS/PS labeling procedure, based on the "click-reaction" of terminal alkynes to azides. While peptide 6 is a β-turn mimic, the all β³-peptide 7 contains a random sequence and is supposed to be linear. Peptide 8, finally, is a helix-mimetic.

Figure 5 PS/PS labeling of peptides 6-8 via on-bead click chemistry Reaction sequence for conducting the on-bead PS/PS labeling reaction of β-peptides on the HMBA linker. Fully deprotected resin samples or individual beads are placed in conical reaction vessels. The samples are treated with an excess of in-situ generated copper (I) and azide modified dye. Irrespective of the actual compound structure, a covalent triazole linkage between the compound of interest and the dye is formed. The final deprotection step yields fluorescence-tagged PS/PS labeled peptide.

Figure 6 Design of a β-peptide library and single bead labeling reaction scheme Design and single bead labeling reaction scheme of the first β-peptide library on the PS/PS platform. The chemical setup consists of an HMBA linker, a propargyl-glycine and a β- alanine spacer, followed by four combinatorial positions. These combinatorial positions are varied within the library by incorporation of different amino acid building blocks. A simple single bead labeling protocol was established for PS/PS labeling of library beads. A typical derivatized bead on the bottom of a conical glass vial is depicted. Figure 7 2D-FIDA anisotropy assay principle for PS/PS-labeled hit compounds 2D-FIDA anisotropy measures rotational motions of molecules diffusing through the confoca! excitation volume. Two detection channels (one vertical, one parallel) are used in combination with a polarized excitation light source. Upon diffusion of TMR-labeled compounds through the excitation volume, the linear polarization of the laser light source leads to a selective excitation of fluorophores with parallel absorption transition dipole moments (photoselection). The degree of depolarization of the excited population, which occurs during the fluorescence lifetime of the fluorophore, depends on the rotational motions of the compound (on the rotational correlation time). The faster the rotation, the higher the depolarization. The photons, emitted during a passage through the excitation volume, are distributed onto the two detectors via a polarizing beam splitter. FIDA histograms and hence apparent molecular brightnesses are calculated for each detection channel. The slower rotation of complexed compounds as compared to free unbound compounds thereby results in lower depolarization, higher intensities in the parallel channel and hence in higher anisotropy values.

Figure 8 FCS assay principle for PS/PS-labeled hit compounds

The detection parameter of fluorescence correlation spectroscopy assays is the diffusion time of molecules through the confocal excitation volume. The autocorrelation function is used to correlate single photons with the detection probability of other photons within short time intervals and, in the end, with molecular passages through the excitation volumes. A fit of the autocorrelation function finally yields the diffusion times associated with the passages of molecules through the confocal volume. The diffusion times of protein complexed molecules are significantly longer as compared to the faster passages of unbound compounds.

Figure 9 The PS/PS labeling method as a bead-based synthesis platform for an integrated screening process EXAMPLES

As used hereafter, the abbreviation PS/PS labeling, refers to "post-synthesis/post-screening" labeling step according to the screening method of the invention.

1. Materials and methods

General methods

HPLC: Vydac peptide C8, 4.6 mm x 150 mm, 5 μm particle diameter size

Gradient: from 5% to 45% ACN in 20 min, then to 95% in 5 min.

PS/PS labeling and analytical methods

Single bead PS/PS labeling and cleavage procedures

Two sample-compartment formats were used for single bead labeling: standard autosampler glass vials with a conical inlet (8002-SC-H/i3 μ, Glastechnik Grafenroda), or 96-well filter plates (polypropylene filter plates with conical shape, 220 μl, GF 5.0 μm, long drip from innovative microplates.com). Single beads were either deposited manually with the aid of a syringe needle (for method development or quality control purposes) or automatically sorted (on the COPAS beadsorter during the screening procedure) into the two types of sample compartments.

Prior to labeling, the bead-containing compartments were filled with methanol (Merck, Uvasol p.a.) and centrifuged for 5 min in a standard SpeedVac vacuum centrifuge. In the case of glass vials, the correct position of beads at the bottom of the conical insert was checked under a stereomicroscope (WILD Heerbrugg).

Labeling a) Glass vials: beads were treated with 12 μl of a four-component labeling solution (4.5 μl H2O, 4 μl tButanol, 1.5 μl catalyst solution, 2 μl dye solution) and agitated for at least 16 h at room temperature by mild shaking. Under the microscope, the labeling solution was removed and the labeled beads were washed with methanol (5x) and water (1x), using a standard micropipette (Gilson, 200 μl). For better efficiency, after each addition of washing solution the beads were allowed to soak and incubate for a few minutes (> 3 min) in methanol. b) Filter plates: After sealing the filter plates at the bottom with a laboratory film ("Parafilm M"), each well was treated with 26 μl of a four-component labeling solution (10 μl H20, 10 μl tButanol, 3 μl catalyst solution, 3 μl dye solution) and finally sealed. The reaction was allowed to proceed under constant agitation for at least 16 hours at room temperature on a microtiter plate shaking device. After removal of the two sealing films (top and bottom) the wells were drained through the filters, employing a commercially available vacuum manifold (VWR International). The labeled beads were then washed by repetitive filling with methanol (5 to 10 times) and draining through the filter. After a final washing step with water, the washed beads were inspected under the microscope and manually transferred into autosampler glass vials using a micropipette (Gilson, Microman M10) with flexible tips. Dye solution: 2 mM methanolic solution of an azide functionalized fluorescent dye Catalyst solution: freshly prepared mixture (1 :1) of ascorbic acid (10 mg/ ml) and copper sulfate (5 mg/ml) in water.

Cleavage

Labeled beads were treated with an ice-cooled solution (6 μl) of NaOH (1 M) / dioxane (1 :1) for 15 min at room temperature. After neutralization with HCI (4 μl, 1 M), the cleavage solution was evaporated under reduced pressure in a SpeedVac vacuum centrifuge. PS/PS batch labeling of beads and HPLC purification to obtain pure, labeled material The single bead labeling protocol employs a high excess of reactive dye (40-400 equivalents) to ensure quantitative compound labeling. As the absolute amounts on one bead are so minute, this excess can be achieved very easily. For resource efficiency reasons and because the crude material was further purified by HPLC, a lower excess (2-10 equivalents) was used for batch labeling of larger compound amounts. Approximately 100 beads were deposited manually into a standard autosampler glass vial with conical inlet, using a syringe needle. The beads were then treated with a four- component labeling solution (15 μl H20, 10 μl tButanol, 5 μl catalyst solution, 10 μl dye solution) and the reaction was allowed to proceed for 24 hours at room temperature under constant agitation. After removal of the labeling solution, the beads were washed and cleaved, essentially as described hereafter.

If a higher number of beads were used for labeling, the volumina of the catalyst and dye solutions were adjusted proportionally to the values given above. After evaporation of the cleavage solution, the crude material was dissolved in water (containing up to 20% of acetonitrile) and purified by HPLC chromatography to yield purified, PS/PS-labeled compounds. HPLC analysis and quantification of single-bead-labeled compounds

The dried material, obtained from a single bead after PS/PS labeling, was dissolved in 20 μl water (containing 5% of acetonitrile) to give a stock solution for further experiments.

For quantification of compound amounts from single beads, 15 μl were injected into the

HPLC. Absolute quantification was achieved by integration of total peak areas in the fluorescence trace (excitation/emission at 555/575 nm) and comparison to a calibration curve

(see below).

For quality control of (hit) compounds, 2 μl (= 1/10 of total material, 1-10 pmoles) of the 20 μl stock solution were diluted into 18 μl of water (containing 5% acetonitrile) and 15 μl thereof analyzed by HPLC.

Calibration curve: Standard solutions of 0.5 to 5 μM TMR azide 5 in water (containing 5% acetonitrile) were prepared and accurate concentrations determined by absorbance measurements of each sample at 555 nm, using an extinction coefficient of 57,100 M^"1 cm^"1.

Integrated peak areas in the fluorescence trace (excitation/emission at 555/575 nm) were fitted to a linear equation with a correlation coefficient R of 0.9988.

Determination of resin loading via Fmoc quantification

Generally, resin loadings were determined after the first synthetic step by measuring the absorbance of the dibenzofulvene-piperidine adduct: two aliquots of the Fmoc-amino acid resin (~ 2 mg) were weighed precisely and suspended in 1 ml piperidine solution (20% in DMF). After 30 min the mixtures were diluted 1:10 and the absorbance was measured at 290 nm on an Agilent 8453 Spectrophotometer in the single beam mode with 50 μl UV quartz microcuvettes and using piperidine solution (20% in DMF) as a reference. The resin loading was calculated according to equation [E-1].

L = ^₂₉₀ l{m x 5.84) _{[E 1 ]}

L: loading [mmol g-1], A290: absorbance at 290 nm, m: amount of resin used [mg].

Spectroscopic methods

Determination of extinction coefficients for TMR-COOH and TMR azide Extinction coefficients were determined from three independent measurements: A defined amount (0.5 to 1 mg) of HPCL-purified TMR azide or commercially available TMR carboxylic acid (Molecular Probes) was weighed and dissolved in water (2 ml). After dilution of 1 :1 ,000 in water (to yield an A550 in the range of 0.1 to 0.01 ) the absorbance of each sample at 550 nm was measured on an Agilent 8453 Spectrophotometer in the single beam mode with 50 μl UV quartz microcuvettes and using H₂O as a reference. Extinction coefficients for each sample were then calculated according to the Lambert-Beer law and averaged over three measurements.

Estimation of solubility of TMR azide 5 in water

Different amounts of TMR azide 5 (1 mg, 0.5 mg, 0.1 mg) were weighed and dissolved in falcon tubes (15 ml) with 5 ml of water by sonication for 10 min. The tubes were then centrifuged for 5 min at 12,000 rpm with a fixed angle SS-34 rotor and classified as dissolved or undissolved by visual inspection. The solubility of TMR azide 5 in water was found to be below 100 μg/ml.

Spectral characterization of free TMR azide 5 and thazole conjugated TMR after click reaction

Solutions of TMR carboxylic acid, TMR azide 5 and TMR PS/PS-labeled peptide 7 in PBS, pH 7.4 (c ~ 0.5 μM) were prepared and filled into quartz cuvettes with 10x10 mm path length (Hellma). For each sample excitation and emission spectra were recorded at 25° C on a SPEX Fluorolog τ-3 spectrofluorometer (Jobin Yvon), using the steady state measurement mode. High voltages of the PMTs were set to 950 V and integration times of 1 s with slit widths of 2/2/2 were used. Excitation spectra were recorded at a resolution of 1 nm for each sample from 350 to 590 nm, with the detection wavelength being 600 nm. Emission spectra were recorded from 535 to 700 nm at a resolution of 1 nm and an excitation wavelength of 520 nm. To obtain corrected excitation and emission spectra, the signal S (measured in counts per second) was further processed as follows:

I _em - Sc / R j_E_₂j h_x = Sc I Rc _[E__3]

U_m> '_ex: corrected emission and excitation intensities, resp.; Sc accounts for background subtraction of a specified blank file and multiplication with a correction for the instrument response. R corrects for lamp intensity fluctuations (reference channel); Rc finally represents the correction file for the non-ideal behavior of gratings and detector.

Determination of quantum yields

Quantum yields were determined following the comparative method of (Williams, Winfield and Miller 1983), where the samples are compared to a standard dye of known quantum yield. Solutions with 4 to 5 different concentrations of rhodamine 6 G, TMR azide 5, TMR carboxylic acid and TMR PS/PS-labeled peptide 7 in PBS were prepared and their OD500 values measured on an Agilent 8453 Spectrophotometer in the single beam mode with 50 μl UV quartz microcuvettes and using PBS as a reference. Then, for each sample a corrected emission spectrum was recorded on the Fluorolog τ-3 spectrofluorometer (Jobin Yvon), as described above at 500 nm excitation and with 1 s integration time and slit widths of 2/2/2. The calculation of the area under the curve for the fully corrected emission spectra yielded the integrated fluorescence intensity for each sample. These integrated fluorescence intensities for each species at different OD500 were fitted to comply with a linear regression of the form y = a^*x +b. The obtained correlation coefficients for the linear regressions ranged from 0.9958 (TMR carboxylic acid) to 0.9999 (TMR PS/PS-labeled peptide 7) and, together with the common interception point, indicated the validity of the measured data. Finally, taking the literature value of 0.95 for the quantum yield of rhodamine 6 G (Kubin and Fletcher 1982), quantum yields for all species were calculated according to equation [E-30].

Determination of fluorescence lifetimes

Frequency domain lifetime measurements

The decay response intensity l(t) of an ensemble of excited molecules with individual lifetimes X₁ can be described as

[E-4]

The real and imaginary components R and I of an impulse response, according to the Fourier transformation are: R = j 1 (t) cos ωtdt o [E-5]

/ = j 1 (t) sin ωtdt

0 [E-6] oxcircular frequency. Substitution of l(t) according to equation [E-4] and simplification gives

j = γ ^ai^Ti^2(Di i 1 + ^T₁ ² _[E__8]

Assuming a monoexponential decay (artj=fi=1 with fi = mole fractions), equations [E-7] and [E-8] are simplified to

/ = Σ ^τω

^{1 + ωχ2} [E-10]

A sinusoidal continuous-wave excitation of the fluorophores in a sample leads to a demodulated response signal with constant phase shift φ and altered signal amplitude M, as compared to the excitation wave.

Phase shift φ and amplitude M are related to the real and imaginary part of the Fourier transform

φ - arctan(/ I R) [Ε-λ λ]

Substitution of equations [E-9] and [E-10] into [E-11] and [E-12] gives φ = arctan ωτ ΓE-131 M¹ =

1 + ω²τ² _[E.₁₄]

Finally, the expressions for the phase lifetime Tφand the modulation lifetime TM become

Frequency domain-based fluorescence lifetime measurements of TMR carboxylic acid and the "clicked" thazole-conjugated form of TMR in peptide 7 were performed on a Fluorolog τ -3 spectrofluorometer (Jobin Yvon), equipped with a Pockels cell and additional hard- and software components required for lifetime measurements.

Using Ludox®, a light scattering standard as reference, phase angles φ and modulation values m were alternately recorded for the two sample solutions and the reference at 13 frequencies from 20 to 150 MHz. Excitation and emission wavelengths of 550/580 nm and 360/360 nm were used for the samples and the reference, respectively. Further instrumental settings were 7/0.5/7 for excitation slits, 7/6/6 for emission slits with PMT high voltages set to 1 ,200 (S), 490 (R), 800 (T). The resulting data sets were fitted to a mono-exponential decay model, using the instrument^'s integrated lifetime modeling software.

Single molecule fluorescence methods - confocal fluctuation spectroscopy

Fluorescence Correlation Spectroscopy (FCS)

FCS measurements were performed on the PS02 instrument, which is based on an Olympus

IX70 inverted microscope and equipped with two avalanche photodiode detectors. A HeNe laser (λ = 543 nm, laser power - 500 μW) served as excitation light source. For TMR detection, a 560DRLP dichroic filter, an interference barrier filter with OD 5 (to block the excitation laser light from the optical detection path) and a 590DF60 bandpass filter were used in the optical path.

For multidimensional analysis, an additional polarization filter in the excitation path and a polarization beamsplitter in the fluorescence emission path were used and the FCS analysis performed in the parallel-polarized detection channel.

All measurements were carried out at ambient temperature (constant at 23.5° C) in 384- or

1 ,536-well glass bottom microtiter plates. A low nM solution of TMR carboxylic acid in assay buffer (adjustment solution) was used to adjust the optical fibers and the confocai pinhole (40 or 70 μm) until a proper count rate (~ 20-50 cpp) and axis ratio (usually between 5 and 10) was observed. The confocai volume was positioned 150 μm above the glass bottom and at least 10 measurements of the adjustment solution (10 s measurement time each) were recorded to determine the actual confocai volume geometries. After this adjustment procedure, fluctuation signals for all samples were recorded in series of at least 10x10 s. Translational diffusion coefficients τD and particle numbers N were obtained by data fitting, using the FIDA Analyze software package (Evotec, version 1.1): The axis ratio of the confocai volume was determined from the adjustment measurement and fixed during the data fitting procedure, whereas the other parameters - triplet fraction, triplet time, translational diffusion time τD, particle number N in the confocai volume and number of diffusing species (components) - were fitted and averaged over > 10 consecutive measurements.

2D-FIDA anisotropy measurements

2D-FIDA measurements were performed on the PS02 instrument, which is based on an Olympus IX70 inverted microscope and equipped with two avalanche photodiode detectors. A HeNe laser (λ = 543 nm, laser power ~ 500 μW) served as excitation light source in combination with a polarization beamsplitter in the fluorescence emission path and an additional linear polarization filter in the excitation path. For TMR detection, a 560DRLP dichroic filter, an interference barrier filter with OD 5 (to block the excitation laser light from the optical detection path) and a 590DF60 bandpass filter were used in the optical path. All measurements were carried out at ambient temperature (constant at 23.5° C) in 384- or 1 ,536-well glass bottom microtiter plates. A low nM solution of TMR carboxylic acid in assay buffer (adjustment solution) was used to adjust the optical fibers and the confocai pinhole (40 or 70 μm) until a proper countrate (~ 20-50 cpp), axis ratio (usually between 5 and 10) and proper FIDA "volume parameters" (a1 = 2e-2, a2 = -2e-3, a3 = 1e4) were observed for each of the two polarization channels. The confocai volume was positioned 150 μm above the glass bottom and at least 10 measurements of the adjustment solution (10 s measurement time each, binning time of 40 μs) were recorded to determine the actual volume parameters and the G-factor. After this adjustment procedure, fluctuation signals for all samples were recorded in series of at least 10x10 s. The molecular brightnesses q for each channel were obtained by data fitting, using the FIDA Analyze software package (Evotec, version 1.1 ): The confocai volume parameters (AO and A1) and the channel-specific brightnesses of the adjustment solution were determined from the adjustment measurement. The G-factor was calculated according to equation [E-17] prior to data fitting of the samples. Finally, the anisotropies were calculated according to equation [E-18] and averaged over > 10 consecutive measurements.

G = 01 * (1 - P true)

<l2 * (1 + P true) [E-17] , ₌ ft ^{~ G} * 42 q_x + 2 * G * q₂ [E-18]

r: anisotropy, qi, q₂: molecular brightnesses in the parallel and perpendicular polarization channel, G: G-factor of the instrument, p_true: experimentally determined polarization (for TMR p_true = 0.034)

Kd determination and quenching analysis

In the simplest case of a ligand/receptor interaction with a 1 :1 stoichiometry, the dissociation constant K_d is defined as

_{Kα =} W M l^RL] [E-19]

[R], [L]: free receptor and ligand concentrations; [RL]: complex concentration at equilibrium. For a (PS/PS-labeled) sample containing the labeled ligand and a percentage of (non- binding competent) impurities fi, the mass balances are

[R] = R₀ - [RL] _{[E 20]}

[L] = L₀ -[RL]- fi * L₀ [E-21]

R₀, L₀: total receptor and ligand concentrations. and the fraction of bound ligand a, derived from the algebraic solution for the binding equation, is α₌ ~ b - WF- 4(R₀L₀ - R_of_imL_Q)

^2L0 _[E-22] with b = (-L₀ - R₀ - K_{d +} J_11nL₀) [E-23] Any measured average steady state readout parameter y (i.e. anisotropy or translational diffusion time) is related to the degree of complex formation by

_ (min+ (max* Q - min) * a) y⁼ I - d - Q) * a _[E._24] with s-\ Hbound qf^ree [E-25] min, max: starting and end values of the average steady state parameter; Q: quenching factor; q_bou_nd, qt_ree: molecular brightnesses of the labeled ligand in the bound and free state. For FIDA measurements, molecular brightnesses are obtained for each polarization channel and the molecular brightness values q can be calculated directly by q = q|| + 2x ql. For FCS measurements, a possible quenching factor can be derived from the total intensities. Equilibrium dissociation constants (K_d values) were obtained by performing a non-linear least square regression fit of the fluctuation data sets, based on equation [E-24] with the software package GraFit 5.0. Equation [E-24] contains two fit parameters: the end-value max and the dissociation constant K_d.

Chemical syntheses

Synthesis of an indole based acid labile safety catch linker

[5-Benzyloxy-1-(toluene-4-sulfonyl)-1 H-indol-3-ylmethyl]-methyl-carbamic acid tert-butyl ester 1

A solution of methylamine in THF (1 M, 17 ml, 12.33 mmol) under cooling on ice was slowly added to a solution of 1-Tosyl- 5-Benzyloxyindole-3-carboxaldehyde (5 g, 12.33 mmol) in CH₂CI₂/acetic acid (10:1 , 500 ml). After stirring for 1 h at room temperature, (polystyrylmethyl)trimethylammonium cyanoborohydride (15 g, 49.22 mmol) was added, and the resulting suspension was agitated for 16 h. The reaction mixture was then filtered and evaporated with repetitive

additions of toluene. The crude product was then dissolved in toluene (20 ml) and treated with di-tert.-butyldicarbonate (8.1 g, 37.0 mmol), followed by triethylamine (17 ml, 123.3 mmol) and stirred for 3 h at room temperature. Column chromatography (SiO₂, ethyl acetate/cyclohexane, 2:8) gave the title compound as a white solid (3.8 g, 59%).

R_f 0.43 [ethyl acetat/cyclohexane (2:8)].

¹H-NMR (500 MHz, CDCI₃): δ= 1.49 (s, 9H, -(CHa)₃), 2.32 (s, 3H, -CH₃), 2.71 (s, 3H, N-CH₃),

4.48 (s, 2H, 1 ^'-H), 5.04 (s, 2H, 1 ^'"-H), 7.01 (pd, 1 H, ³J = 8.8 Hz, 6-H)₁ 7.20 (d, 2H, ³J = 8.3

Hz, 3^"-H, 5^"-H), 7.28 (ps, 1 H, 4-H), 7.32 (t, 1 H, ³J = 7.2 Hz, 4^'"-H), 7.38 (t, 2H, ³J = 7.67

Hz, 3^{" '}-H, 5^'"-H), 7.43 (pd, 3H, 2^"'-H, 6^"'-H, 2-H)₁ 7.72 (d, 2H, ³J = 8.3 Hz, 2^"-H, 6^"-H),

7.87 (d, 1H₁ ³J = 7.5 Hz, 7-H).

¹³C-NMR(125 MHz): 21.53 (-CH₃), 28.47 (-(CHa)₃), 33.09 (N-CH₃), 41.98 (C-1 ^'), 42.86 (C-1 ^'),

70.35 (C-1 ^'"), 70.68 (C-V), 79.68 (C-3^'), 103.66 (C-4), 114.59 (C-7), 115.10 (C-6), 119.44

(C-3), 125.47 (C-2), 126.71 (C-2^", C-6^"), 127.59 (C-2^'", C-6^'"), 127.93 (C-4^'"), 128.52 (C-

3^'", C-5^'"), 129.82 (C-3^", C-5^"), 130.30 (C-7a), 130.88 (C-3a), 135.18 (C-1 ^"), 136.98 (C-

7^'"), 144.84 (C-4^"), 155.72 (C-5).

MS (70 eV, El):

Positive mode, m/z (%) = 1058.3 (30) [2M+NH₄]⁺, 543.1 (10) [M+Na]\ 538.1 (95) [M+NH₄]⁺,

521.1 (30) [M+H]⁺, 465.1 (65) [M-C(CH₃)₃+2H]⁺, 421.1 (10) [M-Boc+2H]⁺, 390.0 (100) [M-

N(CH₃)BoC]⁺.

[δ-Hydroxy-i-^oluene^-sulfonyO-I H-indol-S-ylmethyll-methyl-carbamic acid tert-butyl ester 2

A solution of tosylindole 1 (2.8 g, 5.38 mmol) in THF/EtOH (1 :10, 50 ml) was treated with palladium on activated charcoal

(280 mg, 10% m/m) after flushing the apparatus several times with argon. The flask was then equipped with a H₂ balloon and the reaction allowed to proceed for 16 h under stirring at room temperature. The suspension was then filtered over cellite and dried in vacuo to give a white solid (2.16 g, 93%).

R_f 0.52 [ethyl acetate/cyclohexane (4:6)]. ¹H-NMR (500 MHz, CDCI₃): δ= 1.49 (s, 9H, -(CH₃)₃), 2.31 (s, 3H, -CH₃), 2.70 (s, 3H, N-CH₃),

4.41 (s, 2H, 1 ^'-H)₁ 6.86 (dd, 1H, ³J = 8.8 Hz, ⁴J = 2.2 Hz, 6-H), 7.08 (ps, 1 H, 4-H), 7.18 (d,

2H, ³J = 8.3 Hz, 3"-H, 5^"-H), 7.38 (s, 1 H, 2-H), 7.69 (d, 2H, ³J = 8.3 Hz, 2^"-H, 6^"-H), 7.78

(d, 1H₁ ³J = 8.1 Hz, 7-H).

¹³C-NMR(125 MHz): 21.05 (-CH₃), 28.41 (-(CH₃J₃), 33.20 (N-CH₃), 41.98 (C-1 ^'), 42.87 (C-V),

80.13 (C-3^'), 105.27 (C-4), 114.07 (C-6), 114.62 (C-7), 119.02 (C-3), 125.73 (C-2), 126.68

(C-2^", C-6^"), 129.79 (C-3^", C-5^"), 130.06 (C-7a), 131.09 (C-3a), 135.09 (C-1 ^"), 144.83 (C-

4^"), 152.77 (C-5), 156.10 (C-2^').

MS (70 eV, El):

Positive mode, m/z (%) = 878.3 (35) [2M+NH₄]⁺, 453.0 (10) [M+Naf, 448.1 (100) [M+NH₄]⁺,

431.0 (40) [M+H]\ 375.0 (95) [M-CCH₃+2H]⁺, 300.1 (45), 288.1 (10).

Negative mode, m/z (%) = 859.3 (80) [2M-H]^", 475.1 (30) [M+HCOO]^', 429.1 (100) [M-H]^'.

^-[(tert-Butoxycarbonyl-methyl-aminoJ-methyπ-i-^oluene^-sulfonyO-IH-indol-δ-yloxyl-acetic acid allyl ester 3

Potassium carbonate (481 mg, 3.48 mmol), allylchloroacetate (404 μl, 3.84 mmol) and potassium iodide (156 mg, 0.93 mmol) were added to a solution of hydroxyindole 2 (1 g, 2.32 mmol) in acetone (15 ml), allylchloroacetate (404 μl, 3.84 mmol) and potassium iodide (156 mg, 0.93 mmol). The resulting suspension was stirred at room temperature under nitrogen for 4 days. The reaction mixture was then diluted with water,

the pH adjusted to 6-7 with HCI (1 M) and the solution extracted with CH₂CI₂ (3x50 ml). The combined organic phases were dried (MgSO₄) and evaporated. Purification by column chromatography (SiO₂, ethyl acetate/cyclohexane/methanol, 3.4:6.5:0.1) yielded a colorless oil (1.01 g, 82%).

R_f 0.60 [ethyl acetate/cyclohexane (4:6)]. ¹H-NMR (500 MHz, CDCI₃): δ= 1.49 (s, 9H, -(CHs)₃), 2.34 (s, 3H, -CH₃), 2.70 (s, 3H, N-CH₃), 4.45 (s, 2H, 1 ^'-H), 4.63 (s, 2 H, 1 ^{' "}-H), 4.70 (pd, 2H, ³J=5.7 Hz, 2^{' "}-H); 5.24 (dd, 1 H, ³J = 9.3 Hz, ⁴J = 1.1 Hz, 4^{' "}-H), 5.33 (dd, 1H, ³J = 15.9 Hz, ⁴J = 1.3 Hz, 4^'"-H), 5.88-5.96 (m, 1 H, 3'"-H), 6.99 (pd, 1H, ³J = 8.4 Hz, 6-H)₁ 7.15 (ps, 1 H, 4-H), 7.20 (d, 2H, ³J = 8.3 Hz, 3^"-H, 5"-H), 7.43 (s, 1H, 2-H)₁ 7.71 (d, 2H, ³J = 8.3 Hz, 2^"-H₁ 6^"-H), 7.86 (m, 1H, 7-H). 1³C-NMR(125 MHz): 21.53 (-CH₃), 28.44 (-(CHa)₃), 33.15 (N-CH₃), 42.75 (C-1 ^'), 43.63 (C-V), 65.76 (C-2^'")_τ 66.11 (C-1 ^{' "}), 79.74 (C-3^'), 103.94 (C-4), 114.69 (C-6, C-7), 118.89 (C-4^'"), 119.37 (C-3), 125.63 (C-2), 126.71 (C-2^", C-6^"), 129.60 (C-7a), 129.85 (C-3^", C-5^"), 130.82 (C-3a), 131.48 (C-3^{' "}), 135.10 (C-1 ^"), 144.94 (C-4^"), 154.70 (C-5), 155.90 (C-2^'), 168.53 (COOR).

[3-(([2-(9H-Fluoren-9-ylmethoxycarbonylamino)-pent-4-ynoyl]-methyl-amino)-methyl)-1- (toluene-4-sulfonyl)-1H-indol-5-yloxy]-acetic acid allyl ester 4 Tosylindole 3 (2.07g, 3.92 mmol) was treated with an ice-cold solution of TFA in CH₂CI₂ (50%, 10 ml) for 1 h at room temperature. The reaction mixture was evaporated and the crude product re- dissolved in CH₂CI₂ (15 ml). To this solution HATU (1.8 g, 1.25 mmol), Fmoc- L-propargylglycine (1.71 g, 5.1 mmol)

and N-ethyldiisopropylamine (2 ml, 11.76 mmol) were added at 0° C. The reaction was allowed to warm to room temperature over 16 h under stirring. Then sat. NaHCO₃ (2 ml) was added, the reaction mixture stirred for 10 min and extracted with CH₂CI₂ (3x15 ml). The combined organic phases were washed with brine, dried (MgSO₄) and evaporated in vacuo. Column chromatography (SiO₂, ethyl acetate/cyclohexane/CH₂CI₂, 3.8:6:0.2) yielded the title compound as a white solid (2.9 g, 85%).

Rf 0.33 [ethyl acetat/cyclohexane/CH₂CI₂ (3.8:6:0.2)].

¹H-NMR (500 MHz, DMSO): δ= 2.27 (s, 3H, -CH₃), 2.48-2.6 (m, 2H, 4^'-H), 2.72 (bt, 1H, 6^'- H), 2.79 (s, 3H, N-CH₃), 2.93 (s, 3H, N-CH₃), 4.1-4.8 (m, 2H, 8^'-H), 4.10-4.15 (m, 1H, 9^'-H), 4.18 (t, 1H, ³J = 7.0 Hz, 9^'-H), 4.22 (m, 2H, 8^'-H), 4.49-4.58 (m, 2H, 1 ^'-H), 4.58-4.62 (m, 3H, 3^'-H, 2^"-H), 4.65-4.72 (m, 3H, V-H, 3^'-H), 4.73 (s, 2H, 1 ^{' "}-H), 4.79 (s, 2H, 1 ^'"-H), 5.15 (dd, 1H, ²J = 10.6 Hz, ³J = 1.1 Hz, 4'"-H), 5.27 (dd, 1H, ²J = 15.9 Hz, ³J = 2.3 Hz, 4^'"-H), 5.87 (m, 1H, 3^{' "}-H), 6.95 (dd, ³J = 9.1 Hz, ²J = 2.4 Hz, 6-H), 7.09 (bd, 1 H, ³J = 2.2 Hz, 4-H), 7.24- 7.34 (m, 4H, 16^'-H, 3^"-H, 5^"-H), 7.38 (m, 2H, 12^'-H, 15^'-H), 7.62-7.69 (m, 2H, 10^'-H, 17^'-

H), 7.71 (s, 1H, 2-H)₁ 7.74-7.78 (m, 3H, 7-H, 2^"-H, 6^"-H), 7.83-7.90 (m, 3H, 13^'-H_t 14^'-H,

NH).

¹³C-NMR(125 MHz): 21.05 (-CH₃), 21.34 (C-4'), 33,47 (N-CH₃) 34.21 (N-CH₃), 41.68 (C-V),

44.23 (C-r), 46.69 (C-9^'), 49.89 (C-3^'), 64.95 (C-2^"), 65.05 (C-1 ^'"), 65.85 (C-8^'), 72.66 (C-

6^'), 80.69 (C-5^'), 104.46 (C-4), 113.92 (C-6), 114.16 (C-7), 118.14 (C-4^'"), 118.84 (C-3),

120.14 (C-13\ C-14^'), 125.31 (C-10^', C-17^'), 126.53 (C-2), 126.61 (C-2^", C-6^"), 127.11 (C-

11 ^', C-16^'), 127.68 (C-12^', C-15^'), 129.71 (C-7a), 130.21 (C-3^", C-5^"), 130.63 (C-3a),

132.28 (C-3^{' "}), 134.02 (C-1 ^"), 140.77 (C-13^'a, C-13^'b), 143.80 (C-9^'a, C-17^'a), 145.43 (C-

2^", C-5^"), 154.41 (C-5), 155.80 (C-7^'), 168.44 (COOR), 170.17 (C-2^').

MS (70 eV, El):

Positive mode, m/z (%) = 768.1 (20) [M+Na]⁺, 763.0 (40) [M+NH₄]⁺, 746.1 (100) [M+H]⁺,

398.0 (40).

Negative mode, m/z (%) = 790.5 (10) [M+HCOOV, 566.1 (100).

[3-(([2-(9H-Fluorene-9-ylmethoxycarbonylamino)- pent-4-ynoyl]-methyl-amino)-methyl)-1-(toluene-4- sulfonyl)-1 H-indol-5-yloxy]-acetic acid 5

A solution of allylester 4 (800 mg, 1.07 mmol) in

THF (3 ml) was added to a suspension of polymer bound tetrakis-(triphenylphosphine) palladium (3.3 g, 0.2 mmol) in THF (7 ml) under argon. The resulting suspension was then treated with a

methanolic solution (10 ml) of p-toluene-sulfinate

(275 mg, 1.61 mmol) and stirred at room temperature for 3 h. The reaction mixture was filtered, diluted with water/HCI (20 ml, final pH 2-3) and extracted with CH₂CI₂ (3x15ml). The combined organic phases were dried (MgSO₄) and evaporated in vacuo. Final purification by column chromatography (SiO₂, ethyl acetate/cyclohexane/acetic acid, 6:4:0.05) gave white crystals (1.16 g, 75%).

R_f 0.35 [ethyl acetat/cyclohexane/acetic acid (6:4:0.05)].

¹H-NMR (500 MHz, MeOD): δ= 2.22 (t, 1H, ⁴J= 2.4 Hz, 6'-H), 2.25 (s, 3H, -CH₃), 2.32 (bt, 1H,

6^'-H), 2.55 (ddd, 1 H, ²J= 16.5 Hz, ³J= 5.2 Hz, ⁴J= 2.5 Hz, 4^'-H), 2.65 (ddd, 1H, ²J= 16.5 Hz,

³J= 5.2 Hz, ⁴J= 2.5 Hz, 4^'-H), 2.90 (s, 3H, N-CH₃), 2.99 (s, 3 H, N-CH₃), 4.16 (t, 1 H, ³J= 6.8

Hz, 9^'-Hz), 4.31 (dd, 1 H, ²J= 10.6 Hz, ³J= 6.9 Hz, 8^'-H), 4.37 (dd, 1 H, ²J= 10.6 Hz, ³J= 6.9

Hz, 8^'-H), 4.58 (d, 1H, ²J= 15.8 Hz, 1 ^'-H), 4.59 (s, 2 H, 1 ^{' "}-H), 4.68 (m, 1H, 1^'-H), 4.70 (d,

1H₁ ²J= 15.8 Hz, 1 ^'-H), 4.75 (m, 1H, 3^'-H), 4.81-4.91 (m, 2H, V-H, 3^'-H), 6.99 (dd, 1 H, ³J=

9.2 Hz, ⁴J= 2.3 Hz, 6-H)₁ 7.07 (d, 1H, ⁴J = 2.3 Hz, 4-H), 7.18-7.27 (m, 4-H, 11 ^'-H, 16^'-H, 3^"-

H, 5^"-H), 7.29-7.37 (m, 2H, 12 -H, 15^'-H), 7.57-7.63 (m, 2H, 1O^'-H, 17^'-H), 7.65 (s, 1 H, 2-

H), 7.71-7.78 (m, 4H, 13^'-H, 14^'-H, 2^"-H, 6^"-H), 7.85 (d, 1-H, ³J= 9.2 Hz, 7-H).

¹³C-NMR(125 MHz): 21.44 (-CH₃), 22.65 (C-4^'), 34,48 (N-CH₃), 35.11 (N-CH₃), 43.29 (C-1 ^'),

46.27 (C-1 '),48.38 (C-9^'), 51.41 (C-3^'), 66.52 (C-1 ^'"), 68.02 (C-8^'), 72.12 (C-6^'), 80.35 (C-

5^'), 105.16 (C-4), 115.51 (C-6), 115.79 (C-7), 119.93 (C-3), 120.88 (C-13^', C-14^'), 126.27

(C-10\ C-17^'), 127.93 (C-2^", C-10^"), 128.04 (C-2), 128.13 (C-11 \ C-27^'), 128.74 (C-12\ C-

15^'), 130.94 (C-3^", C-5^"), 131.97 (C-7a), 132.09 (C-3a), 136.14 (C-1 ^"), 142.57 (C-13^'a, C-

13^'b), 145.16 (C-9^'a, C-17^'a), 146.68 (C-4^"), 156.38 (C-5), 158.05 (NCOO), 172.67 (C-2^'),

172.81 (COOH).

MS (70 eV, El):

Positive mode, m/z (%) = 728.1 (20) [M+Na]\ 723.2 (20) [M+NH₄]\ 706.1 (100) [M+H]⁺,

358.0 (40).

Negative mode, m/z (%) = 704.2 (30) [M-H]^", 548.0 (10), 482.1.

([5-Benzyloxy-1-(toluene-4-sulfonyl)-1 H-indol-3-yl]-methanol 6 A stirred solution of i-Tosyl-δ-Benzyloxyindole-S-carboxaldehyde (2.5 g, 6.17 mmol) in dry THF (10 ml) was cooled to 0° C₁ treated with sodium borhydride (332 mmol, 6.78 mmol) and the reaction allowed to proceed for 2 h at room temperature. The solvent was then removed under reduced pressure, the resulting residue re- dissolved in H₂O and neutralized by addition of sat. NH₄CI. After extraction with ethyl acetate (3x15 ml), the combined organic

layers were dried (MgSO₄) and evaporated in vacuo to give the title compound as a white solid (2.46 g, 98%). R_f 0.42 [ethyl acetat/cyclohexane (6:4)].

¹H-NMR (500 MHz, DMSO): δ= 2.29 (s, 3H, -(CH₃)₃), 4.53 (d, 2H, 4J =-CH₃), 2.59 (t, 2H, ³J = 7.1 Hz, 6^'-H), 3.17 (m, 3H, 4^'-H, 7^'-H), ³J = 5.5 Hz, 1 ^'-H), 5.06 (s, 2H, 1 ^'"-H), 5.10 (t, 1 H₁ ³J = 5.5 Hz₁ OH), 7.01 (dd, 1 H, ³J = 11.5 Hz, ⁴J = 2.5 Hz, 6-H)₁ 7.21 (d, 1 H, ⁴J = 2.5 Hz, 4-H), 7.30-7.40 (m, 5H, 3^"-H₁ 5^"-H, 3"^'-H, 4^"'-H, 5^{' "}-H), 7.42 (d, 2H, ³J = 7.8 Hz, 2^{" '}-H, 6^'"-H), 7.55 (s, 1H, 2-H), 7.75-7.80 (m, 3H, 7-H, 3^"-H₁ 5^"-H).

¹³C-NMR(125 MHz): 21.36 (-CH₃), 55.38 (C-T)₁ 70.10 (C-T^"), 104.64 (C-4), 114.48 (C-6, C- 7), 124.58 (C-2), 124.60 (C-3a), 126.95 (C-2^", C-6^"), 128.06 (C-2^"\ C-6^'"), 128.18 (C-4^{' "}), 128.78 (C-3'", C-5^'"), 129.78 (C-7a), 130.53 (C-3", C-5"), 131.02 (C-3), 134.61 (C-T'), 147.46 (C-7^'"), 145.63 (C-4^"), 155.40 (C-5). MS (70 eV, El):

Positive mode, m/z (%) = 852.85 (40) [2M+K]⁺, 836.89 (50) [2M+Na]⁺, 445.88 (100) [M+K]⁺, 429.91 (80) [M+Na]⁺.

3-Hydroxymethyl-1-(toiuene-4-sulfonyl)-1H-indol-5-ol 7

A solution of tosylindole 6 (2.4 g, 5.89 mmol) in THF/EtOH (1:10, 20 ml) was treated with palladium on activated charcoal (240 mg, 10% m/m) after flushing the apparatus several times with argon. The flask was then equipped with a H₂ balloon and the reaction allowed to proceed for 16 h under stirring at room temperature. The suspension was then filtered over cellite and dried in vacuo to give a white solid (1.78 g, 96%).

R_f 0.34 [ethyl acetate/cyclohexane (6:4)].

¹H-NMR (500 MHz, DMSO): δ= 2.28 (s, 3H, -(CH₃)₃), 4.49 (s, 2H, 1 ^'-H), 5.10 (bs, 1H, OH),

6.79 (dd, 1H, ³J = 8.8 Hz, ⁴J = 2.2 Hz, 6-H)₁ 6.90 (d, 1 H, ⁴J = 2.2 Hz, 4-H), 7.31 (d, 2H, ³J = 8.1 Hz, 3^"-H, 5^"-H), 7.48 (s, 1 H, 2-H), 7.68 (d, 1H, ³J = 8.9 Hz, 7-H), 7.75 (d, 2H, ³J = 8.1 Hz, 2^"-H, 6^"-H) 9.35 (bs, 1H, OH).

¹³C-NMR(125 MHz): 21.44 (-CH₃), 55.55 (C-1 ^'), 105.45 (C-4), 114.36 (C-6), 114.45 (C-7), 124.34 (C-2), 124.43 (C-3), 127.02 (C-2", C-6"), 128.89 (C-7a), 130.54 (C-3^", C-5^"), 131.38 (C-3a), 134.74 (C-1^"), 145.55 (C-4^"), 154.37 (C-5).

[3-Hydroxymethyl-1-(toluene-4-sulfonyl)-1 H-indol-5-yloxy]-acetic acid allyl ester 8

To a solution of hydroxyindole 7 (1.86 g, 5.86 mmol) in acetone (15 ml) was added potassium carbonate (1.22 g,

8.79 mmol), allylchloroacetate (1.02 ml, 8.79 mmol) and potassium iodide (188 mg, 1.17 mmol). The resulting suspension was stirred at room temperature under nitrogen for 4 days. The reaction mixture was then diluted with water,

the pH adjusted to 6-7 with HCI (1 M) and the solution extracted with CH₂CI₂ (3x50 ml). The combined organic phases were dried (MgSO₄) and evaporated. Purification by column chromatography (SiO₂, ethyl acetate/cyclohexane, 4:6) yielded a colorless oil (1.58 g, 65%).

R_f 0.60 [ethyl acetat/cyclohexane (4:6)].

¹H-NMR (500 MHz, DMSO): δ= 2.29 (s, 3H, -(CH₃)₃), 4.51 (d, 2H, ⁴J = 3.5 Hz, 1 ^'-H), 4.62 (d,

2H, ³J = 5.4 Hz, 2^'"-H), 4.81 (s, 2H, 1 ^'"-H), 5.12 (s, 1H, OH), 5.19 (dd, 1 H, ³J_E = 9.6 Hz, ²J =

1.3 Hz, 4^'"-H), 5.28 (dd, 1H, ³J_Z = 17.2 Hz, ²J = 1.5 Hz, 4^{' "}-H), 5.84-5.94 (ddd, 1H, 3^'"-H),

6.97 (dd, 1 H, ³J = 9.0 Hz, ⁴J = 2.5 Hz, 6-H), 7.11 (d, 1 H, ⁴J = 2.5 Hz, 4-H), 7.35 (d, 2H, ³J =

8.2 Hz, 3^"-H, 5^"-H), 7.56 (s, 1 H, 2-H), 7.78 (m, 3H, 7-H, 2^"-H, 6^"-H).

¹³C-NMR(125 MHz): 21.46 (-CH₃), 55.42 (C-1 ^'), 65.32 (C-2^'"), 65.51 (C-1 ^{' "}), 104.66 (C-4),

114.27 (C-6), 114.55 (C-7), 118.53 (C-4^'"), 124.62 (C-3), 124.77 (C-2), 127.05 (C-2^", C-6^"),

130.11 (C-7a), 130.65 (C-3", C-5"), 130.98 (C-3a), 132.68 (C-3^'"), 134.65 (C-1 "), 145.79

(C-4^"), 154.65 (C-5), 168.94 (COOR).

MS (70 eV, El):

Positive mode, m/z (%) = 853.16 (100) [2M+Na]\ 438.01 (90) [M+Na]\

[S-Prop^-ynylaminomethyl-i-^oluene^-sulfonyO-IH-indol-δ-yloxyl-acetic acid allyl ester 9 To an ice-cooled solution of allylester 8 (350 mg, 0.84 mmol) in dry acetonitrile (5 ml) was added CBr₄ (334 mg, 1.01 mmol) and triphenylphosphine (264 mg, 0.93 mmol). The reaction was allowed to warm to room temperature and stirred for 2 h. TLC analysis (SiO₂, cyclohexane/ethal acetate, 6:4) showed complete

conversion to a product with R_f = 0.8.

To the reaction mixture propargylamine (0.12 ml, 1.68 mmol) was added and the reaction allowed to proceed further for another 5 h. The reaction mixture was diluted with H₂O and extracted with CH₂CI₂ (3x10 ml). The combined organic layers were dried (MgSO₄) and evaporated in vacouo. Purification of the crude product by column chromatography (SiO₂, cyclohexane/ethyl acetate/ammonia, 6:4:0.01) yielded the title compound as colorless oil

(140 mg, 37%).

R_f 0.23 [cyclohexane/ethyl acetate/ammonia (6:4:0.01)]

¹H-NMR (500 MHz, DMSO): δ= 2.29 (s, 3H, -(CH₃J₃), 3.08 (t, 1 H, ⁴J = 2.3 Hz, 4^'-H), 3.26 (d,

2H, ⁴J = 2.3 Hz, 2^'-H), 4.5.77 (s, 2H, 1 ^'-H), 4.62 (d, 2H, ³J = 5.4 Hz, 2^'"-H), 4.82 (s, 2H, 1 ^'"-

H), 5.19 (dd, 1H, ³J_E = 10.5 Hz, ²J = 1.2 Hz, 4^'"-H), 5.28 (dd, 1H, ³J₂ = 17.2 Hz, ²J = 1.5 Hz,

4^'"-H), 5.84-5.94 (ddd, 1H, 3^{' "}-H), 6.96 (dd, 1H, ³J = 9.0 Hz, ⁴J = 2.5 Hz, 6-H), 7.11 (d, 1H,

⁴J = 2.5 Hz, 4-H), 7.33 (d, 2H, ³J = 8.2 Hz, 3"-H, 5"-H), 7.55 (s, 1H, 2-H), 7.78 (m, 3H, 7-H,

2^"-H, 6^"-H).

¹³C-NMR(125 MHz): 21.45 (-CH₃), 37.22 (C-2^')_t 42.61 (C-1 ^'), 65.30 (C-2^'"), 65.53 (C-V^"),

74.30 (C-4^'), 83.21 (C-3^'), 104.48 (C-4), 114.32 (C-6), 114.60 (C-7), 118.52 (C-4^'"), 122.12

(C-3), 1245.46 (C-2), 127.04 (C-2^", C-6^"), 130.09 (C-7a), 130.6559 (C-3^", C-5^"), 131.56

(C-3a), 132.69 (C-3^'"), 134.51 (C-1 ^"), 145.72 (C-4^"), 154.69 (C-5), 168.94 (COOR).

MS (70 eV, El):

Positive mode, m/z (%) = 927.27 (60) [2M+Na]⁺, 905.28 (40) [2M+H]\ 556.10 (40), 475.11

(100) [M+Na]⁺.

[3-(([3-(2-9H-Fluoren-9-yl-acetylamino)-propionyl]-prop-2-ynyl-amino)-methyl)-1-(toluene-4- sulfonyl)-1 H-indol-5-yloxy]-acetic acid allyl ester 10 To an ice cooled solution of tosylindole 9

(120 mg, 0.27 mmol) in CH₂CI₂ (1.5 ml),

HATU (122 mg, 0.32 mmol), Fmoc-β- alanine (87 mg, 0.28 mmol) and N-ethyl diisopropylamine (135 μl, 0.79 mmol) were added at 0° C. The reaction was allowed to warm to room temperature over

16 h under stirring. Then sat. NaHCO₃

(0.5 ml) was added, the reaction mixture

stirred for 10 min and extracted with

CH₂CI₂ (3x10 ml). The combined organic phases were washed with brine, dried (MgSO₄) and evaporated in vacuo. Column chromatography (SiO₂, ethyl acetate/cyclohexane, 5:6) yielded the title compound as a white solid (140 mg, 71%).

R_f 0.55 [ethyl acetate/cyclohexane (5:6)].

¹H-NMR (500 MHz, DMSO): δ= 2.22 (m, 1H, 4^'-H), 2.32 (s, 3H, -(CH₃)₃), 2.63 (m, 2H, 6^'-H),

3.55 (m, 2H, 6^'-H), 3.81 (s, 2H, 2^'-H), 4.20 (m, 1H, 9^'-H), 4.37 (d, 2H, 8^'-H), 4.54 (s, 2H, 1 ^'"-

H), 4.60-4.75 (m, 4H, 1 ^'-H, 2^"'-H, 1 '^"-H), 5.20-5.35 (m, 2H, 4^'"-H), 5.61 (bs, 1H, NH), 5.75

(bs, 1H, NH), 5.80-5.95 (m, 2H, 2^{' "}-H), 6.85 (m, 1H, A-H), 6.95-7.05 (m, 2H, 4-H, 6-H), 7.20

(d, 2H, ³J = 8.2 Hz, 3^"-H, 5^"-H), 7.29 (m, 2H, 11 ^'-H, 16^'-H), 7.38 (pt, 2H, 12^'-H, 15^'-H), 7.44

(s, 1H, 2-H), 7.53 (s, 1 H, 2-H), 7.58 (m, 2H, 10^'-H₁ 17^'-H), 7.70-7.80 (m, 4H, 13^'-H, 14^'-H,

2^"-H, 6^"-H), 7.85-7.95 (m, 1H, 7-H).

¹³C-NMR(125 MHz): 21.95 (-CH₃), 33.81 (C-6^'), 34.01 (C-2^'), 35.92 (C-2^'), 37.11 (C-7^'),

39.16 (C-1 ^'), 42.54 (C-V), 47.68 (C-9^'), 66.33 (C-1 ^"\ C-2^'"), 67.00 (C-8^'), 72.89 (C-4^'),

73.56 (C-4'), 78.01 (C-3'), 103.77 (C-4), 104.15 (C-4), 115.14 (C-7), 115.49 (C-6), 118.17

(C-3), 119.50 (C-4^'"), 120.34 (C-13^', C-14^'), 125.48 (C-2), 126.97 (C-2), 127.18 (C-2^", C-

6"), 127.42 (C-11 ', C-16'), 128.05 (C-12', C-15'), 130.33 (C-3", C-5"), 131.02 (C-3a),

131.88 (C-3^'"), 135.45 (C-1 ^"), 141.70 (C-13^'a, C-13^'b), 144.38 (C-9^'a, C-17^'b), 145.51 (C-

4^"), 155.30 (C-5), 156.93 (NCO), 169.16 (COOR), 171.74 (C-5^').

^-((^-(ΘH-Fluoren-θ-ylmethoxycarbonylamino^propionyO-prop^-ynyl-aminoJ-methyO-i- (toluene-4-sulfonyl)-1H-indol-5-yloxy]-acetic acid 11 A solution of allylester 10 (90 mg, 0.12 mmol) in THF (1 ml) was added to a suspension of polymer-bound tetrakis(triphenylphosphine) palladium (400 mg, 0.024 mmol) in THF (2 ml) under argon. The resulting suspension was then treated with a methanolic solution (0.25 ml) of p-toluene-sulfinate (32 mg, 0.18 mmol) and stirred at room temperature for 3 h. The reaction mixture was filtered, diluted with water/HCI (5 ml, final pH 2-3) and extracted

with CH₂CI₂ (3x10 ml). The combined organic layers were dried (MgSO₄) and evaporated in vacuo. Final purification by column chromatography (SiO₂, ethyl acetate/cyclohexane/acetic acid, 6:5:0.05) gave a white solid

(50 mg, 59%).

R_f 0.27 [ethyl acetate/cyclohexane/acetic acid (6:5:0.05)].

¹H-NMR (500 MHz, DMSO): δ= 2.22 (m, 1H, 4^'-H), 2.32 (s, 3H, -(CHa)₃), 2.63 (m, 2H, 6^'-H),

3.55 (m, 2H, 6^'-H), 3.81 (s, 2H, 2'-H), 4.20 (m, 1H, 9^'-H), 4.37 (d, 2H, 8^'-H), 4.54 (s, 2H, 1 ^'"-

H), 4.60-4.75 (m, 4H, 1 ^'-H, 2^"'-H₁ 1 ^'"-H), 5.20-5.35 (m, 2H, 4^{' "}-H), 5.61 (bs, 1H, NH), 5.75

(bs, 1H, NH), 5.80-5.95 (m, 2H, 2'"-H), 6.85 (m, 1H, 4-H), 6.95-7.05 (m, 2H, 4-H, 6-H), 7.20

(s, 1H, 2-H), 7.53 (s, 1H, 2-H), 7.58 (m, 2H, 10^'-H, 17^'-H), 7.70-7.80 (m, 4H, 13^'-H, 14^'-H,

2^"-H, 6^"-H), 7.85-7.95 (m, 1H, 7-H).

¹³C-NMR(125 MHz): 21.95 (-CH₃), 33.81 (C-6^'), 34.01 (C-2% 35.92 (C-2^'), 37.11 (C-7^'),

39.16 (C-V), 42.54 (C-V), 47.68 (C-9^'), 66.33 (C-V^", C-2^{' "}), 67.00 (C-8^'), 72.89 (C-4^'),

73.56 (C-4^'), 78.01 (C-3^'), 103.77 (C-4), 104.15 (C-4), 115.14 (C-7), 115.49 (C-6), 118.17

(C-3), 119.50 (C-4'"), 120.34 (C-13', C-14'), 125.48 (C-2), 126.97 (C-2), 127.18 (C-2", C-

6"), 127.42 (C-1 V, C-16'), 128.05 (C-12\ C-15'), 130.33 (C-3", C-5"), 131.02 (C-3a),

131.88 (C-3'"), 135.45 (C-V), 141.70 (C-13'a, C-13^'b), 144.38 (C-9'a, C-17'b), 145.51 (C-

4"), 155.30 (C-5), 156.93 (NCO), 169.16 (COOR), 171.74 (C-5').

MS (70 eV, El):

Positive mode, m/z (%) = 793.23 (100), 728.12 (60) [M+Na]⁺, 706.14 (30) [M+H]⁺.

Negative mode, m/z (%) = 704.19 (50) [M-H]^", 482.12 (100) [M-Fmoc]^".

LCMS: 8.7 min, (electrospray +ve) m/z (%) = 728.2 (10) [M+Na]⁺, 706.1 (100), [M+Hf, 358.4

(30); (electrospray -ve) m/z (%) = 1409.0 (50) [2M-H]^", 704.3 (100) [M-H]^", 482.6 (20).

Purity according to HPLC/UV (215 nm): > 99%. Activation of the indole based safety catch linker resins

Method A: A sample (1 bead to several mg) of thoroughly washed indole 9 containing resin was placed into a small standard autosampler glass vial with conical inlet, using a syringe needle. The beads were treated with a solution of tetrabutylammonium fluoride (1 M in THF) for a reaction time of 3 to 16 hours at constant temperature. For activation under elevated temperatures, the samples were left to react in a water bath. Finally, the activation solution was removed and the beads extensively washed (3x DMF, 2x CH₂CI₂, 3x MeOH).

Method B: The same reagents and procedure as described for method A was were used, but for the activation solution, which consisted of CsF (1 M in water)/NaOH (1 M)/dioxane

(1 :3:3).

Method C: The same reagents and procedure as described for method A were used, but for the activation solution, which consisted of NaOH (1 M)/dioxane (1 :1 ).

Method D: The same reagents and procedure as described for method A was were used, but for the activation solution, which consisted of LiOH/thioglycolic acid (1:3) in DMF.

Cleavage of of the indole based safety catch linker resins after activation and PS/PS-labeling A sample (1 bead to several mg) of thoroughly washed and activated indole resin was placed into a small standard autosampler glass vial with conical inlet. The resin was then treated with a solution of TFA/CH₂CI₂/H₂O (8:2:0.1). After a reaction time of 30 min the sample was dried in vacuo.

Standard procedure for Fmoc deprotection

The resin, containing a terminal Fmoc-protected amino acid, was treated with a solution of 20% piperidine in DMF for 10 min. After filtration the resin was washed once with fresh DMF and the same treatment was repeated 3-times. The resin was then filtered off and washed with DMF (4x3 min), DCM (4x3min).

Standard procedure for coupling ofN-Fmoc amino acids and N-Fmoc-β-homo amino acid building blocks

The N-terminally deprotected and pre-swollen resin (swollen in DCM for a few minutes) was treated with a solution of the appropriate N-Fmoc-βhomo- or N-Fmoc-α amino acid (4 equ.), HATU (3.9 equ.) and DIPEA (8 equ.) in DMF. After agitation for 1 h the resin was filtered off, washed once with DMF and the reaction repeated with fresh reagent. The resin was then drained, washed thoroughly with DMF (4x3min) and DCM (4x3 min) and either subjected to Fmoc deprotection for subsequent couplings or dried in vacuo.

N-terminal acetylation

After the last coupling step and final Fmoc deprotection, the resin was treated with a solution of AC₂O (10 equ.) and DIPEA (20 equ.) in DMF. After a reaction time of 1 h at r.t., the resin was drained and again treated with the same amount of fresh reaction mixture for another 1 h. Finally, the resin was washed extensively with DMF (4x3 min), DCM (4x3 min) and dried under h.v. for 12 h.

Side chain deprotection

All the N-Fmoc amino acid building blocks used for peptide synthesis bear acid-labile protecting groups on their side chains (e.g. tBu, Boc, Trt, Pmc). In order to remove these protecting groups, each resin was treated with a solution containing TFA/H₂O/TIS (95:2.5:2.5) at r.t. for 2 h. The resin was then washed with DCM (4x2 min), 10% DIPEA in DCM (4x2 min), DCM (4x2 min) and dried in vacuo to yield the final resin-bound peptide, which was stored at 4° C.

Loading of TentaGel S HMB resin - synthesis of base resins 1a, b

Esterification of Fmoc-propargyl- glycine (Fmoc-Pra-OH) or Fmoc- β³hpropargyl-glycine with TentaGel S

HMB resin was performed according to standard procedures from

Novabiochem. A solution of Fmoc-

1 a (n=0),

Pra-OH/Fmoc-β³-hPra-OH (4 equ., b (n=1)

5.88 g/6.13 g, 17.56 mmol) in dry DCM (25 ml) was treated with MeIm (3 equ., 1.04 ml, 13.17 mmol), followed by MSNT (4 equ., 5.2 g, 17.56 mmol) at r.t. After complete dissolution of the

MSNT, the reaction mixture was added to the pre-swollen (swollen in DCM for 30 min) resin

(1 equ., 15.15 g, 4.39 mmol). The resulting suspension was then agitated for 1 h at r.t.

Subsequently, the resin was drained and again treated with the same amount of reaction mixture for 1 h at r.t. Finally, the resin was filtered and thoroughly washed with DCM (15 ml,

5x2 min). After load determination indicated a complete loading (> 90% as determined with the standard procedure, described above and compared to the vendor^'s specifications), the

Fmoc group of the propargylglycine was removed according to the standard procedure, described above. The N-terminally deprotected resin (1 equiv., 4.39 mmol) was treated with a solution of N-Fmoc-βAla-OH (4 equ., 5.46 g, 17.56 mmol), HATU (3.9 equ., 6.50 g, 17.01 mmol) and DIPEA (8 equ., 6.01 ml, 35.12 mmol) in DMF (25 ml) and mixed for 1 h. After 1 h the resin was drained, washed once with DMF and the reaction repeated with fresh reagent.

The resin was then filtered off, thoroughly washed with DMF (25 ml, 4x3min) and DCM (25 ml, 4x3 min) and dried in vacuo to give resins 1a, b.

3-azido-propylamine (3)

A solution of 3-amino-proplybromide hydrobromide (400 mg, 1.83 mmol) and sodium azide (8 equ., 940 mg, 14.61 mmol) in water (5 ml) was heated at 85° C for 16 h.

After cooling of the reaction mixture to 4° C, sodium hydroxide (5.6 equ., ² ^N^. -

400 mg, 10.25 mmol) was added and the solution was extracted with 3 diethylether (3x5 ml). The combined organic layers were washed with saturated brine, dried

(MgSO4) and evaporated to give a colorless oil (145 mg, 80%).

3-azido-propylarnido-tetramethylrhodamine (5)

1 H-NMR (500 MHz, CDCI3): δ = 1.82 (m, 2H, 2^"-H), 3.22 (s, 12H, N-CH3), 3.39 (m, 2H, V-

H), 3.45 (m, 2H, 3^"-H), 6.90 (s, 2H, 4-H, 5-H)₁ 7.01 (m, 4H, 1-H, 2-H, 7-H, 8-H), 7.55 (d, 1H,

3J= 7.9 Hz, 6^'-H), 8.28 (dd, 1 H, ³J= 7.9 Hz, 4J= 1.6 Hz, 5^'-H), 8.66 (s, 1H, 3^'-H). 8.93 (t, 1 H,

³J= 5.6 Hz, NH).

13C-NMR(125 MHz): 28.39 (C-2^"), 36.94 (C-1 ^"), 40.50 (NCH3), 48.62 (C-3^"), 96.49 (C-4,

C-5), 114.20 (C-2; C-7), 129.15 (C-3'), 130.00(C-6'), 130.39

(C-1 , C-8), 131.13 (C-2^'), 136.07 (C-4^'), 156.64 (C-3, C-6),

164.82 (CO), 166.22 (COOH).

MS (70 eV, El); m/z (%): 513.27 (100) [M+H]+, 431.19 (12).

2. Results

Schematically, one embodiment of the general chemical setup for the method of the invention can be summarized in Figure 1. It consists of a linkage system called "linker" for attaching compounds to the beads, a covalent core comprising at least one labeling functionality "F", which might be protected during the synthesis and screening steps and, optionally, a "spacer" for tethering the compounds in an appropriate distance to the labeling site or fluorophore. Following a first strategy, after synthesis and screening, any protective group has to be removed prior to performing a labeling reaction on-bead. For this strategy the protection scheme requires three-fold orthogonality as side chain deprotection, labeling functionality deprotection and linker cleavage take place at three different time points. C: Following a second strategy, it employs a polymer supported reagent approach for scavenging any unreacted dye molecules.

A test system and basic resin loading

The peptidomimetics field has been revolutionized in recent years by the discovery that short oligomers of β-amino acids (so called β-peptides) can fold into a variety of stable secondary structures, similar to those found with α-peptides (e.g. helices, sheets or turns). In contrast to α-peptides, these oligomers are stable to proteolytic degradation and therefore hold great unexploited potential as leads and tool compounds (for a review on β-peptides see (Seebach, Beck and Bierbaum 2004)).

As the synthetic procedures for β-peptides are identical with those used for α-peptides and therefore obtainable with equal ease, β-peptides were chosen as compound class for describing one embodiment of the method of the invention. As beta-amino acids are difficult and costly to synthesize, on-bead screening is the only reasonably cheap and fast possibility to assess the interaction potential of important secondary structural elements like β-turns with a full deck of side chains. This approach should therefore allow the identification of target selective β-peptidic high affinity binders from large combinatorial libraries for pharmacological testing.

Hydroxymethylbenzoic acid [HMBA] has been described as one of the most versatile linker systems for SPPS₁ featuring a two-step cleavage procedure (Sheppard and Williams 1982). Side chain deprotection is generally carried out on the support with appropriate amounts of TFA and additives. The resin ester bond, which is stable to acidic conditions, can be cleaved with a variety of nucleophiles to generate C-terminally modified peptides (e.g. acids, amides, esters, alcohols, ...). The possibility to generate resin bound, fully deprotected peptides made this linker an appropriate choice for the purpose of the invention (Figure 2). By combining the HMBA linker with a terminal alkyne-containing amino acid and a β-alanine spacer, a chemical setup with ail necessary elements for PS/PS labeling and the method of the invention was generated:

In a first step, commercially available TentaGel-HMBA resin was loaded with either Fmoc-α- propargylglycine, or Fmoc-β-propargylglycine, using the MSNT, methylimidazole esterification procedure (Blankemeyermenge, Nimtz and Frank 1990). The subsequent load determination via Fmoc-fluorescence showed 100% coupling efficiency after two couplings. As a spacer moiety, one β-alanine was then attached to the resin with HATU as coupling reagent, to give basic resins 1 a,b. For the subsequent peptide syntheses resin 1 served as starting point.

Preparation of Tetramethylrhodamine-azide 5 (TAMRA-azide)

So far, only the resin-bound components for PS/PS labeling were considered. As a second component, a labeling reaction via the 1.3-dipolar cycloaddition reaction to terminal alkynes needs an azide-modified dye. Currently, none of the "standard dyes suitable for confocal fluorescence fluctuation analysis is commercially available in an azido form. We therefore prepared an azide-conjugated tetramethylrhodamine via synthesis of azidopropylamine 3 (Carboni, Benalil and Vaultier 1993) and its condensation with the commercially available tetramethylrhodamine succinimidyl ester 4 (Figure 3). HPLC purification afforded the new TMR-labeling reagent 5 (see Figure 3).

For PS/PS labeling experiments, three β-peptides were synthesized on resin 1 (Figure 4): a mixed α-and β³- tetrapeptide 6 (structure not shown), another all-β³-tetrapeptide 7 (structure not shown), and deca-β³-peptide 8 (structure not shown). The syntheses were carried out applying standard Fmoc solid phase protocols (Chan and White 2000), with HATU as coupling reagent and piperidine for Fmoc deprotection. Side chain deprotection with TFA yielded the final, resin bound peptides 9-11 (Figure 5).

We then used these resin-bound peptides for the assessment of our PS/PS labeling strategy (Figure 5). Initial experiments were based on previously reported reaction conditions (Wu, et al 2004) but led us to the conclusion that for efficient on-bead labeling higher amounts of catalyst (at least a moderate molar excess, relative to the resin-bound alkyne) and labeling reagent (> 10 equivalents) are necessary. For a quantification of labeling yields, approximately 200 beads from resins 9a-c were manually placed into an MS-vial and incubated with 10 equivalents of TMR azide 5 in an aqueous solution, containing copper sulfate and ascorbic acid. After a reaction time of 16 h and several washing steps, the labeled peptides were cleaved from resin by treatment with sodium hydroxide/dioxane for 15 min and subsequent neutralization with HCI (1 M). HPLC analysis of the three samples and comparison to the chromatograms of unlabeled peptides 6-8 showed an almost complete conversion into labeled peptides 11a-c (data not shown). Calculated reaction yields ranged from 85% for peptide 11 c to 90% for 11a, b. The washing steps (3x methanol, 1x water) allowed the efficient removal of excess labeling reagent, leaving no unreacted TMR-azide, according to the HPLC-chromatograms. Additionally, the reaction and cleavage conditions proved compatible with rhodamine dyes, such as TMR. Repeating the cleavage procedure did not result in any further product, indicating complete cleavage after the first 15 min of treatment with sodium hydroxide.

The next step on the route to establishing a generic PS/PS labeling method comprised the downscaling of the click reaction to the single bead level. A first β-peptide library containing more than 80,000 turn-mimetic peptides was synthesized on the HMBA/propargylglycine setup of resin 1a for on-bead screening and PS/PS labeling (Figure 6). Single TentaGel beads from this library were manually placed into conical MS-vials. 90 μm sized TentaGel beads generally have a capacity of 0.2 to 0.3 mmoles/g, which means an average amount of 90-100 pmoles per bead. Each bead was treated over night with 12 μl labeling solution of 2 μl methanolic TMR azide 5 (100 equivalents), 1 μl catalyst solution (copper sulfate/ascorbic acid 1 :1 , 3 mg/ml solution in water) and 9 μl water/tert-butanol (1 :1). After 16 hours the beads turned dark red. A standard microscope facilitated the subsequent washing steps and prevented loss or damage of beads. HPLC analysis after cleavage with 6 μl sodium hydroxide/dioxane (1 :1), followed by neutralization with 4 μl hydrochloric acid, demonstrated successful single bead labeling reactions (data not shown). All HPLC analyzed samples contained a major peak in the TMR fluorescence channel (excitation/emission: 555/575 nm). This indicates a good reproducibility and robustness of the labeling procedure, irrespective of variations in the actual amino acid sequence. Again, the samples contained no trace of unreacted TMR-azide 5.

Quantitative analysis of single bead labeling and cleavage efficiencies Solid phase combinatorial chemistry normally deals with statistically large ensembles, e.g. resin loadings are determined in mmol per gram of bead material. For analysis of single bead reactions, significant variations have to be expected because of the inherent size distributions of the bead material and the dependence of the volume on the third power of the bead radius, r. For a quantitative analysis of our single bead labeling and cleavage procedure, we generated a calibration curve with TMR azide 5, using an HPLC with fluorescence detection at 555 nm excitation and 575 nm emission.

Absolute amounts from two sets of beads were quantified with this method: 41 random beads from the tetra-peptide library (1) and 13 beads which had undergone the screening procedure. The amounts of labeled material cleaved from single beads ranged from 10 to 120 pmoles, 54 and 59 pmoles being the average amounts for the library beads and the screened beads, respectively. These values are in good agreement with the theoretically calculated 50 to 100 pmoles for 90 μm TentaGel beads with a loading of 0.2-0.3 mmoles/g. In addition to the identical average quantities of labeled material, the two sets showed similar bead-to-bead variations. This indicates that no material was lost during the screening procedure. Only over several weeks of storage in methanolic or aqueous solution significant cleavage of peptides on the beads occurred, resulting in yields as low as 2 to 10 pmoles per bead.

The concentration of TMR labeled li-peptides after single bead labeling was determined by FCS. This concentration determination is based on the assumption that the confocal volume can be approximated by a cylinder with height 2 xω2 and diameter 2 x ω.1. Both, height and diameter are accessible via the axis ratio and the measured diffusion time τD of a standard dye (TMR-COOH in this case) according to Equations [E-26] and [E-27]:

2

CO. τD = AD [E-26]

ω = AR * fl>

2 1 [E-27] where D denotes the translational diffusion coefficient D and for TMR equals 2.8 x 10.₆ [crτi₂ S_-1]. Inserting the average particle number N, determined by FCS, into equation [E-28] finally gives the concentration of fluorescent molecules in the sample as:

N

C =

N_A * fi> ²π * 2«

^{1 2} [E-28]

NA: Avogadro number, 6.023x1023, c: concentration [mol L₁].

Absolute amounts determined with this method for 10 cleaved fractions from single bead labelings ranged from 20 to 60 pmoles. A comparison to the respective amounts determined by HPLC shows that FCS measurements result in slightly lower values for the same solution.

Nevertheless, a clear overall correlation with a correlation coefficient (R2 = 0.905) was obtained.

Spectroscopic characterization of PS/PS-TMR labeled compounds

2D-FIDA-anisotropy and FCS - two confirmation methods for labeled hit compounds Modern fluorescence-based single molecule detection (SMD) methods provide a powerful toolbox for measuring biomolecular interactions in microliter-sized sample volumes (Scheel, et al 2001). The repertoire of available techniques ranges from fluctuation analysis via auto- or cross correlation (FCS, FCCS) via intensity-based histogramming methods (1 D-FIDA, 2D- FIDA-anisotropy, 2D-FIDA-2-color) to lifetime-based (cFLA, cTRA) methods (Gall, et al 2002; Kask, et al 1999; Kask, et al 2000; Palo, et al 2002). Coupled analysis techniques (FIMDA, FILDA) further increase and stabilize the possible detection modes extractable from a single measurement (Gall, et al 2002; Palo, et al 2002; Palo, et al 2000a; Palo, et a! 2000b). However, the more sophisticated methods suffer from longer measurement times and higher complexity. This holds especially true for the single burst analyses methods, like BIFL, also called MDF (Eggeling, et al 1998; Eggeling, et al 2005; Fries, et al 1998). As hit compounds from OBS are exclusively small with molecular weights lower than 1 ,000 Da, the two classical techniques used for quantification of target protein binding are FCS and 2D-FIDA which detect changes in the translational diffusion time t^ and the fluorescence anisotropy r of a labeled molecular species, respectively.

One of the practical limitations of FCS and also (to a lesser extent) 2D-FIDA measurements in resolving biomolecular interactions is the need for a substantial difference in the reaction partners' molecular masses - or, more accurately, in their translational (FCS) or rotational (2D-FIDA) diffusion times (Meseth, et al 1999).

The solution confirmation of PS/PS labeled hit compounds represents the nearly perfect biochemical system for the FCS and 2D-FIDA methods, as the mass difference between the labeled ligand (the compound) and the target protein is usually at least 4- to 5-fold or higher. Hence, these two measurement techniques were used for off-bead confirmations (Figures 7 and 8). Expected values for diffusion times W and anisotropies r of PS/PS derivatized OBS hit compounds

On-bead binding and direct solution confirmation after single bead labeling To test the feasibility of a direct single bead-based solution confirmation, on-bead and subsequent solution binding experiments were carried out using peptides 15-18. We tested the on-bead binding of a biotinylated target to the β-peptides under four different blocking conditions: 0.5% BSA, 50 μM E. coli lysate, StartingBlock blocking buffer from Pierce and 3 mg/ml gelatine. Using 50 nM target and 10 nM red emitting Qdots655-streptavidin, all four peptides 15-18 exhibited target binding on-bead with no significant differences concerning signal intensity and background. This indicates that even under stringent conditions like the presence of a 1 , 000-fold excess of E. coli proteins, a micromolar binder will still show up in the on-bead assay. However, dye-labeled target (Cy5 or Alexa488) did not result in any binding on-bead.

Single beads from the well with peptide 15 under lysate blocking were then manually removed and subjected to PS/PS labeling. One half of the beads was HPLC-purified prior to single point measurements with target, the other half was directly measured without purification. With both methods, 2D-FIDA and FCS, the single bead material without purification resulted in significantly lower end values upon protein addition as compared to the purified peptide 15 (data not shown). This difference arises from by-products and impurities, originating from the synthesis. Normally, on-bead compounds from a library synthesis comprising 5 to 10 synthetic steps reach a purity of 60 to 90%, the main impurities usually being shorter molecules with one or more building block deletions. These side products most likely will not bind to the target and therefore constitute a second species in the measurement, which contributes to the signal of the unbound fraction. Furthermore, this second species, heterogeneous in size, can be expected to reduce the start and end values. The crude substance derivatized from peptide 15 is only 40% pure according to the HPLC chromatogram. Therefore this sample represents a "worst" case with respect to the detection of binding events in non-purified single bead-derived compounds. The limited resolution of the individual FCS and 2D-FIDA assays obtained from unpurified single bead cleavage fractions can be significantly improved by combining the 2D-FIDA and FCS readouts into one 2D-analysis. This also simplifies the detection of false positives: for a real hit the translational and rotational diffusion data must always stay roughly on the diagonal, while false positives would show a significant deviation on either of the two axes. Furthermore, the use of a smaller pinhole for the fluctuation measurements also increases the resolution in the FCS mode. With the use of the 40 μm pinhole the translational diffusion time for the peptide 15 sample increases by 42% upon target protein binding as compared to 16% increase with the 70 μm pinhole.

(Re)-identification of a β-peptidic target binder within a large library - multiplexing in on-bead screening

In addition to providing a direct off-bead confirmation possibility, the PS/PS labeling method opens up new routes for "multiplexing" in on-bead screening. Previous examples of target- multiplexed on-bead screens involved the use of differentially labeled target proteins (Dixon, et al 2005). In these approaches the targets needed to be detected by spectral separation. This in general limits the number of targets mixed into one screen to a few. In reality, only examples for screening two targets in one screen can be found in the literature. A further possible complication arises from the influence of different dye chemistries used for target labeling on the compound-binding properties on-bead. Even perfectly selective on-bead hits may turn out to be "dye-compromised" false positives, meaning that the selectivity might be due to an adverse effect of the other label on binding to the compound. Therefore the best strategy is using the same dye for every target in a mixed protein composition within one experiment.

The PS/PS technology allows for a new breath of multiplexing - the screening of on-bead chemical space against a biological space, spanned by a (theoretically unlimited) number of uniformly labeled targets. The on-bead screen is conducted with a mixture of uniformly labeled targets. It ideally reduces the chemical space to its (target-) relevant size, represented by a series of primary on-bead hits. At this stage hits are only known to bind to one or more targets within the mixture. For these hits, their respective target and a specificity profile are assigned through solution testing of every compound against each target. The PS/PS labeling methodology makes this deconvolution step possible and overcomes the need for spectral separation of the individual targets during the on-bead screen.

Methods for the retrieval of hit beads

Currently, two methods for the retrieval of hit beads are presented hereafter: (a) The PickoScreen instruments contain a picking device which uses the xy-coordinates from the CONA scan to position a thin glass capillary accurately above a hit bead. In a semi- automated fashion the hit bead is then soaked up by the glass capillary and dispensed into a standard MS-vial. Picking on the PS04 instrument also includes optical feedback, wich allows to re-check the image of the selected beads immediately before picking. This picking procedure takes approximately 2 min per bead and is the method of choice if only a few hits or a specific set of beads are to be picked.

(b) The second method for hit bead retrieval uses a COPAS beadsorter (Union Biometrica), capable of analyzing combinatorial library beads at a speed of 20 to 30 beads per second. Our COPAS is specifically equipped with a high frequency profiling device and records 100 to 200 intensity values as a bead passes through the flow cell. These intensity values result in a bead profile rather than a mere "integrated fluorescence signal" and allow the elimination of false positives, based on the fluorescence profile: e.g. if a bead shows high local fluorescence intensities due to target precipitation, it would nevertheless be classified as hit bead in a mere "integrated fluorescence signal"-based method.

3. Discussion

Hit confirmation in a reliable and quantitative way is a central problem in on-bead screening, as well as in any screening process in general. Especially all "heterogeneous" screening methods, such as chip- or bead-based systems, need to translate their primary detection signals into the relevant solution binding information. Starting from the idea of directly linking an on-bead binding event to its confirmation in homogeneous solution, we developed a method for derivatizing compound on single beads after synthesis and screening. In finding an appropriate chemical setup, several requirements had to be met, the most demanding of which was to allow full compound deprotection for screening, followed by site-specific derivatization and cleavage. This "PostSynthesis/PostScreening" labeling method reliably transforms individual hit beads into fluorescently tagged compounds ready for binding and competition studies in solution in cell lysates and in cells. In a first experimental realization of the PS/PS concept, we used a standard HMBA linker in combination with the 1 ,3-dipolar cycloaddition of azides to terminal alkynes, known as click chemistry. The initial setup fulfills all requirements for a generic application in the field of (β)-peptide chemistry.

PS/PS labeling, a bead-based synthesis/derivatization method allowing for on-bead, in vitro and cellular screening

Although the PS/PS labeling methodology was originally developed with the idea to bridge the gap between on-bead binding and solution confirmation, the method holds the potential for an application in a broader lead discovery context. The invaluable advantage of a PS/PS- based synthesis platform is the optional generation of fluorescent ligands in a generic way. Moreover, in a case-by-case choice any tagging group and any dye can potentially be used to tag a given library without deciding beforehand (during the synthesis phase) on a specific tagging group. Thus the PS/PS technology can be used to setup an affinity-based "multimodal" screening process. "Multi-modal" in this context means the simultaneous usage of one chemical library (input) for at least three different screening strategies: a) in vitro nanoscreening b) cellular screening and c) on-bead screening (Figure 9). Each screening system benefits from the advantages of having fluorescent ligands - e.g. the in vitro nanoscreening with these compounds could be carried out using every target without any lengthy assay development.

Taken together, a PS/PS technology-based synthesis and screening process offers an unprecedented degree of flexibility in the field of chemical biology and drug discovery.

References

Alexander MD, Burkart MD, Leonard MS, et al (2006) A central strategy for converting natural products into fluorescent probes. ChemBioChem; 7 (3):409-16.

Alluri PG, Reddy MM, Bachhawat-Sikder K, et al (2003) Isolation of protein ligands from large peptoid libraries. J.Am.Chem.Soα; 125 (46): 13995-4004.

Amini F, Denison C, Lin HJ et al (2003) Using oxidative crosslinking and proximity labeling to quantitatively characterize protein-protein and protein-peptide complexes. Chem.Biol.; 10 (11):1115-27.

Atherton E, Logan CJ, Sheppard RC (1981) Peptide synthesis. Part 2. Procedures for solid- phase synthesis using Nα-fluorenylmethoxycarbonylamino-acids on polyamide supports. Synthesis of substance P and of acyl carrier protein 65-74 decapeptide. J.Chem.Soα, Perkin Trans. 1 , 538-46.

Auer M, Gstach H (2001 ) Inventors. Novartis AG, assignee. Fluorescent dyes (AIDA) for solid phase and solution phase screening. United States Patent, US 6,207,831 B1 , granted Mar 27 2001.

Backes BJ, Harris JL, Leonetti F, et al (2000) Synthesis of positional-scanning libraries of fluorogenic peptide substrates to define the extended substrate specificity of plasmin and thrombin. Nat.Biotechnol.; 18 (2):187-93.

Beebe KD, Wang P, Arabaci G, et al (2000) Determination of the binding specificity of the SH2 domains of protein tyrosine phosphatase SHP-1 through the screening of a combinatorial phosphotyrosyl peptide library. Biochemistry; 39 (43): 13251 -60. Blackwell HE, Perez L, Stavenger RA, et al (2001 ) A one-bead, one-stock solution approach to chemical genetics: part 1. Chem.Biol.; 8 (12):1167-82.

Blankemeyermenge B, Nimtz M, Frank R (1990) An efficient method for anchoring Fmoc- amino acids to hydroxyl-functionalized solid supports. Tetrahedron Lett.; 31 (12):1701-4.

Bradner JE, McPherson OM, Mazitschek R, et al (2006) A robust small-molecule microarray platform for screening cell lysates. Chem.Biol.; 13 (5):493-504.

Carboni B, Benalil A, Vaultier M (1993) Aliphatic amino azides as key building-blocks for efficient polyamine syntheses. J.Org.Chem.; 58 (14):3736-41.

Chan WC, White PD (eds) (2000) Fmoc Solid Phase Peptide Synthesis: A Practical Approach Oxford:Oxford University Press. demons PA, Koehler AN, Wagner BK, et al (2001) A one-bead, one-stock solution approach to chemical genetics: part 2. Chem.Biol.; 8 (12):1183-95.

Costantino L, Barlocco D (2006) Privileged structures as leads in medicinal chemistry. Curr.Med.Chem.; 13 (1):5-85.

Dixon S, Robins L, Elling RA, et al (2005) Discovery of selective aldo-keto reductase ligands - an on-bead assay strategy. Bioorg. Med. Chem. Lett.; 15 (11):2938-42.

Eggeling C, Fries JR, Brand L, et al (1998) Monitoring conformational dynamics of a single molecule by selective fluorescence spectroscopy. Proc.Natl.Acad.Sci. U. S.A.; 95 (4):1556-61.

Eggeling C, Jager S, Winkler D, et al (2005) Comparison of different fluorescence fluctuation methods for their use in FRET assays: Monitoring a protease reaction. Curr.Pharm.Biotechnol.; 6 (5):351-71.

Fries JR, Brand L, Eggeling C, et al (1998) Quantitative identification of different single molecules by selective time-resolved confocal fluorescence spectroscopy. J.Phys.Chem.A; 102 (33):6601-13.

Gademann K, Kimmerlin T, Hoyer D, et al (2001) Peptide folding induces high and selective affinity of a linear and small beta-peptide to the human somatostatin receptor 4. J.Med.Chem.; 44 (15): 2460-8.

Gall K, Palo K, Brand L, et al (2002) Fluorescence intensity and lifetime distribution analysis: A new and highly sensitive fluorescence fluctuation method. Biophys.J.; 82 (1):355A.

Holmes CP, Jones DG (1995) Reagents for combinatorial organic synthesis: Development of a new o-Nitrobenzyl photolabile linker for solid phase synthesis. J.Org.Chem.; 60 (8):2318-9.

Horton DA, Severinsen R, Kofod-Hansen M, et al (2005) A versatile synthetic approach to peptidyl privileged structures using a "safety-catch" linker. J. Comb. Chem.; 7 (3):421-35.

Hu B, Xiong B, Qiu BY, et al (2006) Construction of a small peptide library related to inhibitor OM99-2 and its structure-activity relationship to b-secretase. Acta Pharmacol. Sin.; 27 (12):1586-93. Kask P, Palo K, Fay N, et al (2000) Two-dimensional fluorescence intensity distribution analysis: Theory and applications. Biophys.J.; 78 (4):1703-13.

Kask P, Palo K, Ullmann D et al (1999) Fluorescence-intensity distribution analysis and its application in biomolecular detection technology. Proc.Natl.Acad.Sci. U.S.A.; 96 (24):13756- 61.

Khersonsky SM, Jung DW, Kang TW, et al (2003) Facilitated forward chemical genetics using a tagged triazine library and zebrafish embryo screening. J.Am.Chem.Soc; 125 (39): 11804-5.

Kodadek T, Bachhawat-Sikder K (2006) Optimized protocols for the isolation of specific protein-binding peptides or peptoids from combinatorial libraries displayed on beads. Mol.Biosyst.; 2 (1):25-35.

KoIb HC, Sharpless KB (2003) The growing impact of click chemistry on drug discovery. Drug Discov.Today; 8 (24):1128-37.

Kubin RF, Fletcher AN (1982) Fluorescence quantum yields of some rhodamine dyes. J.Luminesc; 27 (4):455-62.

Lam KS, Lebl M, Krchnak V (1997) The "one-bead-one-compound" combinatorial library method. Chem.Rev.; 97 (2):411-48.

Lam KS, Liu R, Miyamoto S, et al (2003) Applications of one-bead one-compound combinatorial libraries and chemical microarrays in signal transduction research. Acc.Chem.Res.; 36 (6):370-7.

Lam KS, Salmon SE, Hersh EM, et al (1991 ) A new type of synthetic peptide library for identifying ligand-binding activity. Nature; 354 (6348):82-84.

Lehman A, Gholami S, Hahn M, et al (2006) Image subtraction approach to screening one- bead-one-compound combinatorial libraries with complex protein mixtures. J.Comb.Chem.; 8 (4):562-570.

Lin HJ, Kodadek T (2005) Photo-induced oxidative cross-linking as a method to evaluate the specificity of protein-ligand interactions. J.Pept.Res.; 65 (2):221-8.

Liu B, Alluri PG, Yu P, et al (2005) A potent transactivation domain mimic with activity in living cells. J.Am.Chem.Soc; 127 (23):8254-5.

Maclean D, Hale R, Chen M (2001) The reversed Kenner linker: a new safety-catch linker for the preparation of N-alkyl sulfonamides. Org.Lett.; 3 (19):2977-80.

Meisner NC, Hintersteiner M₁ UhI V, Weidemann T, Schmied M, Gstach H₁ Auer M (2004a) The chemical hunt for the identification of drugable targets. Curr.Opin.Chem.Biol.; 8 (4):424- 31.

Meldal M (2002) The one-bead two-compound assay for solid phase screening of combinatorial libraries. Biopolymers; 66 (2):93-100. Meldal M (1992) Pega: a flow stable polyethylene glycol dimethyl acrylamide copolymer for solid phase synthesis. Tetrahedron Lett.; 33 (21 ):3077-80.

Meseth U, Wohland T, Rigler R, et al (1999) Resolution of fluorescence correlation measurements. Biophys.J.; 76 (3):1619-31.

Metzger A, Diller DJ, Lin TH, et al (2006) Successful screening of large encoded combinatorial libraries leading to the discovery of novel p38 MAP kinase inhibitors. Comb.Chem.High Throughput.Screen.; 9 (5):351-8.

Mitsopoulos G, Walsh DP, Chang YT (2004) Tagged library approach to chemical genomics and proteomics. Curr.Opin.Chem.Biol.; 8 (1):26-32.

Muckenschnabel I, Falchetto R, Mayr LM, et al (2004) SpeedScreen: label-free liquid chromatography-mass spectrometry-based high-throughput screening for the discovery of orphan protein ligands. Anal.Biochem.; 324 (2):241-9.

Nunn C Rueping M, Langenegger D, et al (2003) Beta(2)/beta(3)-di- and alpha/beta(3)- tetrapeptide derivatives as potent agonists at somatostatin sst(4) receptors. Naunyn Schmiedebergs Arch. Pharmacol.; 367 (2):95-103.

Palo K, Brand L, Eggeling C, et al (2002c) Fluorescence intensity and lifetime distribution analysis: toward higher accuracy in fluorescence fluctuation spectroscopy. Biophys.J.; 83 (2):605-18.

Palo K, Jager S, Mets U, et al (2000a) FIMDA, a new method which covers both, FCS and FIDA. Biophys.J.; 78 (1):385A.

Palo K, Metz U, Jager S, Kask P, et al (2000c) Fluorescence intensity multiple distributions analysis: Concurrent determination of diffusion times and molecular brightness. Biophys.J.; 79 (6):2858-66.

Patek M, Lebl M (1991) Safety-catch anchoring linkage for synthesis of peptide amides by Boc/Fmoc strategy. Tetrahedron Lett.; 32 (31):3891-4.

Paulick MG, Hart KM, Brinner KM, et al (2006) Cleavable hydrophilic linker for one-bead- one-compound sequencing of oligomer libraries by tandem mass spectrometry. J.Comb.Chem.; 8 (3):417-26.

Plunkett MJ, Ellman JA (1995) A silicon-based linker for traceless solid-phase synthesis. J.Org.Chem.; 60 (19):6006-7.

Rademann J, Grøtli M, Meldal M, et al (1999) SPOCC: A resin for solid-phase organic chemistry and enzymatic reactions on solid phase. J.Am.Chem.Soα; 121 (23):5459-66.

Reddy MM, Bachhawat-Sikder K, Kodadek T (2004) Transformation of low-affinity lead compounds into high-affinity protein capture agents. Chem.Biol.; 11 (8):1127-37.

Renil M, Meldal M (1996) POEPOP and POEPS: Inert polyethylene glycol crosslinked polymeric supports for solid synthesis. Tetrahedron Lett.; 37 (34):6185-8. Rink H (1987) Solid-phase synthesis of protected peptide fragments using a trialkoxy- diphenyl-methylester resin. Tetrahedron Lett.; 28 (33):3787-90.

Rothwell PJ, Berger S, Kensch O, et al (2003) Multiparameter single-molecule fluorescence spectroscopy reveals heterogeneity of HIV-1 reverse transcriptase:primer/template complexes. Proc.Natl.Acad.Sci. U.S.A.; 100 (4):1655-60.

Scheel AA, Funsch B, Busch M, et al (2001) Receptor-ligand interactions studied with homogeneous fluorescence-based assays suitable for miniaturized screening. J.Biomol.Screen.; 6 (1 ):11-8.

Seebach D, Beck AK, Bierbaum DJ (2004) The world of beta- and gamma-peptides comprised of homologated proteinogenic amino acids and other components. Chem.Biodiv.; 1 (8):1111-239.

Sheppard R C, William, BJ (1982) Acid-labile resin linkage agents for use in solid phase peptide synthesis. Int.J.Pept.Prot.Res.; 20 (5):451-4.

Song A, Zhang J, Lebrilla CB, et al (2003) A novel and rapid encoding method based on mass spectrometry for "one-bead-one-compound" small molecule combinatorial libraries. J.Am.Chem.Soc; 125 (20):6180-8.

Sweeney MC, Wavreille AS, Park J, et al (2005) Decoding protein-protein interactions through combinatorial chemistry: sequence specificity of SHP-1 , SHP-2, and SHIP SH2 domains. Biochemistry; 44 (45): 14932-47.

UhI V, Gstach H, Seifert JM, Jung T, Falchetto R, Fetsch A, Stange R, Schmied M, Graf C₁ Adam A, Amberger F, Monteux D, Prechelmacher S, Pruckner A, Dautz W, Graf P, Hermes E, van Hooft P, Berces A, Mϋller J, Auer M (2002) Confocal nanoscanning - bead picking - AIDA technology. A methodology for drug discovery with orphan molecular targets and combinatorial chemistry on the solid surface. Single MoL; 3 (2-3):174-5.

Urbina HD, Debaene F, Jost, B, et al (2006) Self-assembled small-molecule microarrays for protease screening and profiling. ChemBioChem; 7 (11):1790-7.

Uttamchandani M, Walsh DP, Khersonsky SM, et al (2004) Microarrays of tagged combinatorial triazine libraries in the discovery of small-molecule ligands of human IgG. J.Comb.Chem.; 6 (6):862-8.

Wang X₁ Peng L, Liu R, Gill SS, et al (2005a) Partial alloc-deprotection approach for ladder synthesis of "one-bead one-compound" combinatorial libraries. J.Comb.Chem.; 7 (2):197- 209.

Wang X, Peng L, Liu R, et al (2005b) Applications of topological^ segregated bilayer beads in 'one-bead one-compound' combinatorial libraries. J.Pept.Res.; 65 (1):130-8.

Widengren J, Kudryavtsev V, Antonik M, et al (2006) Single-molecule detection and identification of multiple species by multiparameter fluorescence detection. Anal.Chem.; 78 (6):2039-50.

Williams ATR, Winfield SA, Miller JN (1983) Relative fluorescence quantum yields using a computer controlled luminescence spectrometer. Analyst; 108: 1067. Wu P, Feldman AK, Nugent AK, et al (2004) Efficiency and fidelity in a click-chemistry route to triazole dendrimers by the copper(l)-catalyzed ligation of azides and alkynes. Angew.Chem.lnt.Ed.; 43 (30):3928-32.

Ying L, Liu R, Zhang J, et al (2005) A topological^ segregated one-bead-one-compound combinatorial glycopeptide library for identification of lectin ligands. J.Comb.Chem.; 7 (3):372-84.

Youngquist RS, Fuentes GR, Lacey MP, et al (1994) Matrix-assisted laser desorption ionization for rapid determination of the sequences of biologically active peptides isolated from support-bound combinatorial peptide libraries. Rapid Commun.Mass.Spectrom.; 8 (1):77-81.

Zehender H, Le Goff F, Lehmann N, et al (2004) SpeedScreen: the "missing link" between genomics and lead discovery. J. Biomol. Screen.; 9 (6):498-505.

Claims

1. A starting material of formula (I), for on-bead compound library synthesis :

B - L - C - S

(I) wherein

- represents a bond or any stable covalent spacer,

and wherein L and C are selected so that

2. The starting material of Claim 1 , wherein L is a base labile linker or a safety catch linker and C is an alkyne-containing amino acid or a terminal alkyne group as part of the linker.

3. The starting material of Claim 1 , having the following formula (II)

(H) wherein B is a resin bead and R is a protected or unprotected functional group suitable for on-bead compound synthesis.

4. The starting material of Claim 1, having the following formula (III)

wherein B is a resin bead and R is a protected or unprotected functional group suitable for on-bead compound synthesis

5. The starting material of Claim 1 , having the following formula (IV)

wherein B is a resin bead

6. An on-bead compound of formula (V)

B - L - C - X

(V) wherein

- , B, L and C are as defined in any of Claims 1-5, and,

X is any synthetic test molecule.

7. The on-bead compound of Claim 6, wherein L is a base labile linker or a safety catch linker and C is a terminal alkyne group or an alkyne-containing amino acid.

8. The on-bead compound of Claim 6, having the formula (Vl)

(Vl)

wherein B is a resin bead and X is any synthetic test molecule.

9. The on-bead compound of Claim 6, having the formula (VII)

wherein B is a resin bead and X is any synthetic test molecule.

10. The on-bead compound of Claim 6, having the formula (VIII)

X

wherein B is a resin bead and X is any synthetic test molecule.

11. An on-bead compound library comprising a mixture of on-bead compounds of formula (V) as defined in any of Claims 6-10, wherein solid support carries many copies of a single structure species of a synthetic test molecule and said library comprises at least 100 different structure species of synthetic test molecules, preferably at least 10,000, and more preferably one million different structure species of synthetic test molecules.

12. A labeled on-bead compound of formula (IX)

B - L - C - X

I LABEL

(IX)

wherein

- , B, L and C and X are as defined in any of Claims 1-11 , and LABEL is a labeling or detection molecule.

13. The labeled on-bead compound of Claim 12, wherein L is a base labile linker or a safety catch linker and C is an alkyne-containing amino acid or a terminal alkyne group as part of the linker.

14. The labeled on-bead compound of Claim 12, having the formula (XIII)

(XIIl) wherein LABEL is a labeling or detection molecule, B is a resin bead and X is any synthetic test molecule.

15. The labeled on-bead compound of Claim 12, having the formula (XIV)

wherein LABEL is a labeling or detection molecule, B is a resin bead and X is any synthetic test molecule.

16. The labeled on-bead compound of Claim 12, having the formula (XV)

17. The labeled on-bead compound of any of Claims 12-16, wherein the LABEL is selected among the group consisting of

(2) a radioactively labeled molecule;

(3) a molecule which is one part of a specific binding pair capable of specifically binding to a partner, e.g., biotin, which can bind to avidin or streptavidin;

(4) an enzyme capable of catalyzing a detectable reaction;

(5) a molecule possessing a combination of any of the properties listed above.

18. The labeled on-bead compound of Claims 12 or 13, wherein said LABEL is a fluorescent dye.

19. A compound of any of Claims 1 , 6 or 12, wherein said solid support B is selected among the group consisting of a polymer bead, thread, pin, sheet, membrane, silicon wafer, and a grafted polymer unit.

20. A compound of Claim 19, wherein said solid support is made of resin bead.

21. The compound of Claim 20, wherein said resin is selected among the group consisting of: functionalized based resins, polyacrylamide based polymers, PolyOxyEthylene- PolyOxyPropylene polymers, Super Permeable Organic Combinatorial Chemistry polymers, polystyrene/polydimethylacrylamide composites, PEGA resins, polystyrene-polyoxyethylene based supports, Tentagel.

22. A method for identifying a test molecule that has a functional effect upon a target of interest, said method comprising:

(a) contacting on-bead compound libraries as defined in Claim 11 with a target of interest,

(b) determining the functional effect of on-bead compound members upon the target, and

(c) isolating the on-bead compound carrying the test molecule having a functional effect upon the target.

23. The method of claim 22, wherein said functional effect is a target-compound binding event.

24. The method of Claim 23, wherein said target is labeled with a fluorescent or luminescent dye prior to step (a).

25. The method of Claim 24, wherein at step (b), binding of an on-bead compound to the target is determined by confocal nanoscanning.

26. The method of Claim 24 or 25, wherein at step (c) the individual on-bead compound carrying the test compound having binding capacity with the target is isolated by sorting and/or picking with an appropriate device.

27. The method of any of claims 22-26, further comprising after step (c):

(d) deprotecting the labeling functional group of the covalent linking core C of the isolated on- bead compound from step (c),

(e) reacting a labeling molecule with the covalent linking core C of the isolated on-bead compound, thereby yielding a labeled on-bead compound as defined in any of Claims 8-12,

(f) cleaving the linker under appropriate conditions to release the labeled synthetic test compound from the bead,

(g) recovering the labeled synthetic test molecule from step (f).

28. The method of Claim 27, wherein the released labeled test molecule obtained at step (g) is used to confirm its binding capacity to the target in solution.

29. The method of Claim 28, wherein the labeling molecule used at step (e) is a fluorescent dye and binding capacity to the target in solution is confirmed by use of appropriate spectroscopic methods such as conventional (non-confocal) fluorescence anisotropy, 2D- FIDA, FIDA, FCS or FIMDA, BIFL, cTRA, FILDA, cFLA.

30. The method of any of Claims 27-29, wherein a portion of the released labeled test molecule is used to confirm its binding capacity to the target in solution and another portion is used for decoding of structure of the test molecule having binding capacity to the target.

31. The method of any of Claims 27-30, wherein a portion of the released labeled candidate is used in a cell penetration assay or in a cellular assay.